The Use of Virtual Reality Environments in Cognitive Rehabilitation after Traumatic Brain Injury

INTRODUCTION

In the 1990s, the wider dissemination of research outcomes combined with low'er cost computer hardware encouraged researchers in cognitive rehabilitation to develop clinical virtual reality (VR) environments. It was hoped that VR would improve the standardization of traditional neuropsychological testing and enable larger amounts of data to be collected and compared. How'ever, at the same time, one main criticism of traditional neuropsychological testing was that it did not reflect actual cognitive functioning in daily life (Rose, 1996). To address this limit, researchers sought to develop alternative neuropsychological tests which had “functional and predictive relationships” with natural performance (Sbordone, 1996). The concept of ecological validity (EV) appeared to be a key criterion in a new generation of neuropsychological tests aimed both to reflect and predict effects of cognitive disorders in an "open environment” (Franzen & Wilhelm, 1996). Shallice & Burgess’s seminal work (Shallice & Burgess, 1991) concerning EV resulted in the creation of the Multiple Errands Test (MET) which assessed executive functioning in persons after traumatic brain injury (TBI). Indeed, the MET enabled the identification of impairments in daily functioning, in other words dysexecutive syndrome, reported by patients or families, that traditional neuropsychological test batteries could not identify (Burgess et al., 2006). The topic of EV generated a significant debate about the need for more realism and prediction of environmental behavior in neuropsychology testing (Franzen & Wilhelm, 1996).

In this context, it appeared that simulating naturalistic environments and activities using VR may have some advantages that could meet this need (Rizzo et al., 2004). Furthermore, computer systems enabled the development of objective behavioral indicators in these ecologically valid but also safe environments (Schultheis & Rizzo, 2001). Virtual environments could also benefit from experimental control, have multiple demand levels, deliver stimuli and opportunity for rehearsal that led to virtual cognitive rehabilitation (Rose et al., 2005). For Weiss et al. (2006) « the ultimate goal ofVR-based intervention is to enable clients to become more able to participate in their own real environments in as independent manner as possible » (Weiss et al., 2006). Therefore, VR became increasingly more important in a naturalistic-orientated conception of CR, with its focus on the transferability and generalizability of compensation strategies into everyday life (Cicerone et al., 2019). However, although it is easier using VR to simulate realistic scenes, a key point is to structure VR for CR assessment or training on the basis of current cognitive theories.

VR GENERAL PRINCIPLES APPLIED TO CR AFTER TBI

It is challenging to face the complexity of cognitive and behavioral disorders in CR. Within acquired brain injury, TBI is a major concern throughout the world (Maas et al., 2017). Moderate to severe TBI leads to complex clinical situations that combine motor, cognitive, and behavioral disorders that have a major impact in daily functioning (Ponsford et al., 2014). This impact is of great concern, leading to stress, burden, and mood disorders for families (Poulin et al., 2019). For this population with TBI, impacts on activities and participation are related to instrumental activities of daily living (IADL) such as household chores, management of finances, diary management, shopping, and also outside navigation, transportation, and return to work (Dawson and Chipman, 1995). Daily life complexity may be defined in relation to the significant number and diversity of interactions with environmental contexts (Spector et al., 1987). Cognitive demand is notably higher when dealing with multitasking, problemsolving, and coping with environmental needs. Achieving goal-directed activities requires the mobilization of cognitive control resources (Cooper & Shallice, 2000) and to have good self-awareness in order to interact with the environment (Katz & Hartman-Maeir. 2005). In addition, this ability, which mobilizes cognitive resources to plan and organize actions and adapt voluntary behavior to the environment to achieve a goal, is called executive functioning (Lezak, 1982). Dysexecutive Syndrome has a tremendous impact on daily life and long-lasting community reintegration for patients with TBI (Caron et al., 2018).

VR clinical application in CR seeks to capture the activity demands of daily life in order to mobilize cognitive resources. Virtual technology makes it possible to interact in real time through simulated activities and so enables immersion and interaction (Fuchs, 2018). Available scripts can mirror a large variety of IADL (for instance, preparing a meal, doing errands in a mall, driving a car) or tasks included in IADL (e.g., moving around in a town, navigating to find a route). These scripts demand analytical and adaptation skills from the person in order that they can respond to the demands created in VR simulation. For persons with cognitive impairment, it requires them to use control resources and to process information by acting and carrying out several tasks simultaneously. Recording of behavioral data allows the clinician to develop clinical indicators and to use replay with clients to encourage self-awareness and efficiency of cognitive strategies. Virtual activity rehearsal in the same controlled conditions or multicontextual training (Toglia et al., 2011) promotes learning and generalization of efficient behavior (Klinger et al., 2010). Weiss et Col. proposed a cognitive model of the experience of patients using VR in CR (Weiss et al., 2006). They identified three levels of experience: the Interaction Space between the user and VR, where the user experiments and interacts with the virtual environment; the

Transfer Phase which refers to the process that leads to transfer and generalization of skills or strategy to the real-world, and the Real World Environment, where participation is targeted as the ultimate goal of CR. Clinicians who use VR in CR must first identify their CR goals: What has to be improved? What is the main goal of the intervention? How can they meet these goals making the best use of VR, the ultimate aim being to improve the client’s participation in everyday life (Kizony, 2018)?

Many recent literature reviews have explored the usefulness and evidence of effectiveness of the use of VR in CR for persons withTBI (Pietrzaket al., 2014; Shin and Kim, 2015; Imhoff et al., 2016; Aida et al., 2018; Alashram et al., 2019; Banville et al., 2019; Maggio et al., 2019; Manivannan et al., 2019.) They strongly underline the issues mentioned above in relation to this specific population. At the same time, they underline the importance of developing VR to provide clinicians with effective tools that complement other more traditional CR approaches. All authors agree that the use of VR in CR for persons after TBI offers a significant potential in the assessment and treatment of people with TBI, whatever the stage of care, from the initial phase to long-term follow-up. However, despite the promising findings from the existing studies, the reviews underline the need to carry out high-powered and long-lasting follow-up experimental studies particularly in relation to CR and moderate to severe TBI (Imhoff et al., 2016; Alashram et al., 2019; Maggio et al., 2019).

Technological devices used in existing studies are essentially non immersive devices based on the use of both computer or laptop keyboard with mouse or joystick to interact and a computer screen for visual feedback (e.g., Besnard et al. 2016). Visual feedback can be from a video projection on a large screen to create a semi-immersive display (Figure 7.1). Finally, and contrary to common

FIGURE 7.1 Semi-immersive environment in cognitive rehabilitation. (Reprinted with permission - CMRF Kerpape - France.)

representations about VR. few clinical studies use fully immersive devices such as a head mounted display (HMD) (e.g., Dahdah et al., 2017). Despite many technological improvements in recent years (Mayor), there is still concern about the risk of cybersickness when using fully immersive devices, particularly with neurological patients. However, in studies with people with TBI, fewproblems have been reported (Cox et al., 2010; Dahdah et al., 2017; Robitaille et al., 2017).

In persons with moderate to severe TBI, dysexecutive syndrome is compounded when performing tasks that are not familiar (Poncet et al., 2017). Thus, the use of VR could increase task demand if the use of VR devices and virtual environment are not familiar. One way to control the familiarity of the technical use of VR is to begin any VR intervention with a familiarization time. This makes it possible to control both the technological variable and the functional features of the virtual environment, and thus avoid interpretation bias on the quality of performance (Klinger et al.. 2010).

A variety of simulated activities of daily living are used with persons with TBI. For instance, cooking tasks such as preparing a hot beverage (Besnard et al., 2016), or preparing a simple meal (Zhang et al., 2003), or heating up a snack in a microwave oven (Yip & Man, 2009). Other tasks may concern navigating outside to learn or find a route (Caglio et al., 2012; Sorita et al., 2013). Navigation in simulated real environments is surprisingly little reported in VR literature while clinical literature reports significant problems of topographical orientation in TBI patients in their daily life (Boyd and Sautter, 1993). Other studies of VR have enabled a deeper understanding of the underlying cerebral processes involved in topographical disorders among the broad population of people with TBI (Cogné et al., 2017). Driving is a complex activity involving specific motor and cognitive skills. It is a major issue for people with TBI, who have a relatively young mean age. Several studies explore road safety and respect of driving rules in this population (Cox et al., 2010; Milleville-Pennel et al., 2010) or the underlying cognitive processes involved in driving (Lengenfelder et al., 2002). Considering the prevalence of executive disorders among people with TBI, the most important part of tasks in VR involve behavioral planning to reach a goal, time, and space organization and multitasking. Among these kinds of complex adaptative tasks are office task management (McGeorge et al., 2001), library management (Renison et al., 2012), or multiple errands tasks in the supermarket or mall (Jacoby et al., 2013; Canty et al., 2014). Multiple errands tasks in the supermarket are among the most widely used daily living tasks in VR. The supermarket is a natural, complex setting that mobilizes brain resources and notably executive functioning, which is impacted in numerous patients with brain injury (Josman et al., 2008). Finally, in several studies, the authors rely on commercial video games (Grealy et al., 1999; Caglio et al., 2012). The main objective of intervention is to improve memory or attention by stimulating cognitive processes.