Medarević Jelena, Miljković Nadica, Stojmenova Pečečnik Kristina, Sodnik Jaka
Faculty of Electrical Engineering, University of Ljubljana, Ljubljana, Slovenia.
School of Electrical Engineering, University of Belgrade, Belgrade, Serbia.
Front Physiol. 2025 Mar 5;16:1480018. doi: 10.3389/fphys.2025.1480018. eCollection 2025.
Distress detection in virtual reality systems offers a wealth of opportunities to improve user experiences and enhance therapeutic practices by catering to individual physiological and emotional states.
This study evaluates the performance of two wearable devices, the Empatica E4 wristband and the Faros 360, in detecting distress in a motion-controlled interactive virtual reality environment. Subjects were exposed to a baseline measurement and two VR scenes, one non-interactive and one interactive, involving problem-solving and distractors. Heart rate measurements from both devices, including mean heart rate, root mean square of successive differences, and subject-specific thresholds, were utilized to explore distress intensity and frequency.
Both the Faros and E4 sensors adequately captured physiological signals, with Faros demonstrating a higher signal-to-noise ratio and consistency. While correlation coefficients were moderately positive between Faros and E4 data, indicating a linear relationship, small mean absolute error and root mean square error values suggested good agreement in measuring heart rate. Analysis of distress occurrence during the interactive scene revealed that both devices detect more high- and medium-level distress occurrences compared to the non-interactive scene.
Device-specific factors in distress detection were emphasized due to differences in detected distress events between devices.
虚拟现实系统中的压力检测通过迎合个体的生理和情绪状态，为改善用户体验和加强治疗实践提供了大量机会。
本研究评估了两款可穿戴设备，即Empatica E4腕带和Faros 360，在运动控制的交互式虚拟现实环境中检测压力的性能。受试者接受了一次基线测量以及两个虚拟现实场景，一个是非交互式的，另一个是交互式的，涉及解决问题和干扰因素。利用两款设备的心率测量数据，包括平均心率、逐次差分的均方根以及受试者特定阈值，来探究压力强度和频率。
Faros和E4传感器均能充分捕捉生理信号，Faros的信噪比和一致性更高。虽然Faros和E4数据之间的相关系数呈中度正相关，表明存在线性关系，但较小的平均绝对误差和均方根误差值表明在心率测量方面一致性良好。对交互式场景中压力发生情况的分析表明，与非交互式场景相比，两款设备检测到的中高水平压力发生情况更多。
由于设备之间检测到的压力事件存在差异，因此强调了压力检测中特定于设备的因素。
