Human-Computer Co-Regulation: A Review of Biofeedback for Meditative Practice

Abstract

Biofeedback has been proposed to support meditation and mindfulness practices through wearables, mobile applications, VR/AR systems for decades, and other interactive devices. However, evidence of its effectiveness remains mixed and fragmented across existing research in psychophysiology and Human-Computer Interaction (HCI). This work analyzes 39 studies published between 2015 and 2025 on real-time, user-facing biofeedback for contemplative practice. We first outline a taxonomy of biofeedback interfaces based on sensing modality, feedback channel, interaction context and temporal feedback characteristics, then we synthesize the existing study results on modality-practice fit, showing that the performance of each biofeedback sensing modality in meditative practice is closely related to the meditation practice type: Respiration and heart rate variability (HRV) are the most suitable modalities for paced breathing and stress recovery, and EEG is more appropriate for attention-oriented and experiential targets. Electrodermal activity (EDA) mainly reflects arousal awareness rather than directly measuring meditation quality. In addition, a review-informed co-regulation framework consisting of within-session physiological regulation, experiential mediation, and transfer support is proposed in this work. Finally, we summarize the main gaps in the existing literature and propose design principles for future systems.