A Virtual-Reality Approach for the Assessment and Rehabilitation of Multitasking Deficits

Introduction

It has repeatedly been proposed that neurocognitive assessment and rehabilitation should include ecologically valid procedures, i.e., it should be realistic, functionally relevant and as complex as daily life (e.g., Chaytor & Schmitter-Edgecombe, 2003; Schultheis & Rizzo, 2001; Wollesen & Voelcker-Rehage, 2014). At the same time, procedures should allow the examiner to fully control ambient conditions, stimulus presentation and registration of behavioral responses, as is the case in typical laboratory research. Virtual reality (VR) offers the potential for fulling both requirements: it can be designed such as to be ecologically valid and well-controlled (Schultheis & Rizzo, 2001).

VR technology has been applied with success for the assessment and rehabilitation of neuropsychological disorders (for recent reviews, see Howard, 2017; Jovanovski & Zakzanis, 2017), including attention deficits (Rizzo et al., 2000), spatial neglect (Ogourtsova, Souza Silva, Archambault, & Lamontagne, 2015), and upper limb dysfunction (Laver, George, Thomas, Deutsch, & Crotty, 2015). Our study presents a VR approach for the assessment and rehabilitation of multitasking deficits.

The ability to process multiple tasks at the same time is critical for a wide range of everyday activities such as car driving, pedestrian walking, grocery shopping and meal preparation. It is degraded in patients suffering from stroke (Burgess et al., 2006), Parkinson’s disease (Willemsen, Grimbergen, Slabbekoorn, & Bloem, 2000), Alzheimer’s disease (Esposito et al., 2010) or schizophrenia (Laloyaux et al., 2014), as well as in healthy older adults (Verhaeghen, Steitz, Sliwinski, & Cerella, 2003). Assessment and rehabilitation of multitasking skills is typically administered by computer software which displays abstract stimuli on a monitor and registers participants’ responses with a joystick, computer mouse or keyboard. This approach lacks ecological validity for several reasons: stimuli often are abstract rather than realistic, the number of concurrent tasks rarely exceeds two, and participants’ responses are not natural (e.g., participants “walk” through a displayed scene by depressing keys rather than by moving their legs).

VR scenarios for ecologically valid multitasking are already available, mainly for grocery shopping and food preparation (e.g., Rand, Weiss, & Katz, 2009; Zhang et al., 2001). VR scenarios for car driving and pedestrian street crossing have been established as well, but all past approaches were limited to two instead of multiple concurrent tasks: driving and street crossing has been combined either with music listening (Neider et al., 2011) or with phone conversation (Horberry, Anderson, Regan, Triggs, & Brown, 2006; Horrey & Wickens, 2006), mobile internet use (Byington & Schwebel, 2013), text messaging (Drews, Yazdani, Godfrey, Cooper, & Strayer, 2009), cockpit instrument manipulation (Horberry et al., 2006), object detection (Cassavaugh & Kramer, 2009), or arithmetic operations (Chaparro, Wood, & Carberry, 2005). The present study introduces an approach which, for the first time, combines driving and street crossing with a whole battery of realistic concurrent tasks. Preliminary data illustrate that this approach is indeed sensitive to the effects of multitasking on driving and street crossing.