Electrophysiological Correlates for the Detection of Haptic Illusions

Haptic Illusions (HIs) have emerged as a versatile method to enrich haptic experiences for computing systems, especially in virtual reality scenarios. Unlike traditional haptic rendering, HIs do not rely on complex hardware. Instead, HIs leverage multisensory interactions, which can be elicited through audio-visual channels. However, the intensity at which HIs can be effectively applied is highly subject-dependent, and typical measures only estimate generalized boundaries based on small samples. Consequently, resulting techniques compromise the experience for some users and fail to fully exploit an HI for others. We propose adapting HI intensity to the physiological responses of individual users to optimize their haptic experiences. Specifically, we investigate electroencephalographic (EEG) correlates associated with the detection of an HI's manipulations. For this, we integrated EEG with an established psychophysical protocol. Our user study (N = 32) revealed distinct and separable EEG markers between detected and undetected HI manipulations. We identified contrasts in oscillatory activity between the central and parietal, as well as in frontal regions, as reliable markers for detection. Further, we trained machine learning models with simple averaged signals, which demonstrated potential for future in situ HI detection. These discoveries pave the way for adaptive HI systems that tailor elicitation to individual and contextual factors, enabling HIs to produce more convincing and reliable haptic feedback.