Work

An innovative web-based approach to generate respiratory biofeedback using virtual reality headset and embedded inertial sensors during an imposed breathing protocol: a proof-of-concept study

Year: 2025

Type: article

Abstract: Assessing impact of VR immersive scenarios on users, objective biometrics are measured by external sensors, interfering with subject’s behavior and preventing outside laboratory applications. Micro-electromechanical systems embedded in VR headsets allow measuring biomarkers such as respiratory rate by exploiting head-related micro-movements by head-ballistocardiographic (HBCG) technique. We aimed at conducting a proof-of-concept study by developing a device-independent web-based imposed breathing VR experience during which the HBCG signal was acquired. Duration of respiratory cycle (tRESP HBCG ) during such VR experience was estimated in quasi-real time, thus allowing potential use in ecological momentary assessment of emotion elicitation. Twenty-six healthy volunteers were recruited for the VR experience: they were asked, while sitting, to breath at different frequencies (4 s, 6 s, 8 s, 10 s per breath, corresponding to 15, 10, 7.5, and 6 respiratory cycles/ min) following an audio-visual guide. The HBCG signals were recorded and given as input for parameters identification to a grey-box algorithm to optimize the tRESP HBCG estimation, by using as reference the respiratory rate derived from simultaneously acquired 1-lead ECG. The optimized algorithm run on an online server to provide quasi real-time results. It was tested on 8 additional volunteers, reaching an absolute estimation error below 4 %. In addition, robustness analysis results showed good ability of the implemented algorithm in providing accurate results also in presence of noise. The proposed web-based approach resulted feasible and accurate, opening up new opportunities to create VR immersive scenarios and applications to measure biofeedback, with potential use in ecological momentary assessment of emotion elicitation and in respiratory training.

Source: Biomedical Signal Processing and Control

DOI: 10.1016/j.bspc.2025.107966

Authors: Andrea Carpi, Sarah Solbiati, Valentino Megale, Enrico Gianluca Caiani

Institutions: Politecnico di Milano, IRCCS Istituto Auxologico Italiano

Language: English

Topic: Non-Invasive Vital Sign Monitoring

Subfield: Biomedical Engineering

Field: Engineering

Domain: Physical Sciences

SDG: Good health and well-being

Open Access status: hybrid

Estimated APC paid: $2,590