Background Post-stroke attentional deficits are heterogeneous and may involve both spatial attention impairments, such as unilateral spatial neglect, and non-spatial deficits, including distractor inhibition. Conventional cognitive neurorehabilitation tasks often rely on fixed or manually adjusted difficulty levels to match the level of the patient, which may not adequately accommodate individual variability. Dynamic Difficulty Adjustment (DDA) algorithms offer a performance-driven approach to modulating task demands. The present study examined the feasibility of a theory-driven DDA algorithm integrating individual performance and normative data within a non-immersive virtual reality environment task. Methods A single-group repeated-measures feasibility study was conducted in inpatient and outpatient rehabilitation settings. Ten individuals with stroke completed three 15-minute sessions of DDA-REAsmash. The multi-parameter DDA algorithm continuously adjusted stimulus presentation time, distractor inhibition demands, and spatial configuration based on real-time individual performance, aiming to maintain an optimal engagement range of 65–85% success rate. Normative reaction-time percentiles derived from an independent mean age-matched sample were implemented as lower algorithmic constraints. Feasibility was defined as stabilization of performance within predefined engagement boundaries, and adaptive behavior was examined in relation to baseline attentional profiles. Results All participants completed the protocol without adverse events. Across sessions, performance converged toward the target engagement range. Conclusions These findings indicate the feasibility of a theory-driven multi-parameter DDA algorithm integrating individual performance and normative data to maintain individualized engagement in a heterogeneous post-stroke sample. Further controlled studies are required to determine clinical efficacy and functional outcomes.
Sorrentino et al. (2026) studied this question.