The REAsmash serious game for the post-stroke diagnosis of distractor inhibition: contrast between immersive and non-immersive virtual reality test versions.

Background: Virtual reality (VR) Serious Games (SG) offer greater sensitivity and specificity than traditional diagnostics. The playfulness of the SG reduces stress, enhancing motivation and reliability. We developed immersive (iVR) and non-immersive (niVR) versions of REAsmash, a SG based on Feature Integration Theory (FIT) to assess distractor inhibition attention.

Aim: The aim of this study was to verify the transfer of the REAsmash FIT diagnostic properties across VR devices with different degrees of immersion.

Design: Cross-sectional clinical study.

Setting: Inpatient, outpatient and healthy controls.

Population: Post-stroke and healthy individuals.

Methods: The REAsmash involves searching for a (target) mole with a red miner's helmet. The target is either presented alone (baseline), or presented with distractors (11, 17 or 23) that contrast the target by high or low saliency (moles with blue miner's and horned helmets vs blue miner's and red horned helmets). Stimuli appeared randomly from a 24-molehill grid. Participants (15 with and history of cortical-subcortical stroke and 15 age matched controls) hit the target with their response hand in niVR and with a virtual hammer in iVR. Post-stroke participants used their less impaired hand, controls their dominant hand. ANOVA tested VR type (niVR vs iVR), group (post-stroke vs healthy), saliency (high vs low) and distractor number (11, 17, 23), with the interaction between saliency and distractor number defining FIT. The dependent variable was relative mean response time, calculated by subtracting the mean baseline response time from each response to targets presented with distractors, for each participant. This variable exemplifies the costs to response time cause by the manipulation of independent variables.

Results: We found significant main effects and an interaction for saliency and distractor number, confirming FIT. Group and VR type main effects were significant, with slower responses for post-strokes and for iVR, but with no interactions.

Conclusions: To evaluate performance across acute to chronic post-stroke phases, diagnostic measures must be transferable between test devices, ensuring compatibility from hospital to outpatient settings.

Clinical Rehabilitation Impact: Our results demonstrated that the REAsmash diagnostic properties were consistent across immersive and non-immersive VR, as well as within both groups of participants.