| Sumario: | This study investigates the integration of brain-computer interface (BCI) technology with pseudo-haptic feedback to enhance adaptive interactions in virtual reality (VR). Using a NeuroSky headset, we developed a dual-layer control system combining blink detection (via SVM classifier) and attention-level modulation to operate 16 pseudo-haptic button configurations. Manifold learning (UMAP) and wavelet analysis revealed that sustained attention (>70%) dynamically adjusted haptic intensity (23% deeper protrusion) and beta-band (13–30 Hz) energy peaks predicted successful interactions 500 ms pre-trigger. Subjective evaluations showed proximity feedback with protrusion significantly improved embodiment (η² = 0.35), while hit effects boosted satisfaction (p < 0.01). BCI-driven adaptation achieved 82% user satisfaction – validating its necessity for personalised feedback. Results demonstrate BCI's capacity to align cognitive engagement with tactile realism, offering empirical guidelines for adaptive VR interfaces in neurorehabilitation and immersive training. Key metrics: 89.4% blink detection accuracy, 68.2% mean attention, task speed-attention correlation (r = 0.53).
|
|---|