Technical validation of a multimodal emotion-adaptive biofeedback system for autonomic regulation using guided breathing

Breathing interventions delivered through biofeedback have been shown to support physiological modulation consistent with resonance breathing responses. However, most biofeedback systems on the market are based on fixed procedures, provide limited physiological feedback, and do not adapt to user comfort or emotional state. The system is designed to work under a typical 20-min biofeedback timeframe and uses a two-stage adaptive respiration rate (RR) algorithm that gradually converges toward resonance frequency while dynamically responding to signs of emotional discomfort. System-level testing was conducted using reference-grade measurement instruments. Respiration rate estimation showed low mean absolute error, while heart rate and heart rate variability measures demonstrated high agreement with reference standards. During adaptive biofeedback, the system was associated with a decrease in respiratory rate and electrodermal activity, along with increased heart rate variability, consistent with increased parasympathetic influence inferred from HRV trends. These results indicate the technical validity and system-level feasibility of an emotion-aware multimodal biofeedback framework. The proposed framework provides a scalable foundation for future controlled and longitudinal validation studies.