Heart and Breathing Rate Variations as Biomarkers for Anxiety Detection

With advances in portable and wearable devices, it should be possible to analyze and interpret the collected biosignals from those devices to tailor a psychological intervention to help patients. This study focuses on detecting anxiety by using a portable device that collects electrocardiogram (ECG) and respiration (RSP) signals. The feature extraction focused on heart-rate variability (HRV) and breathing-rate variability (BRV). We show that a significant change in these signals occurred between the non-anxiety-induced and anxiety-induced states. The HRV biomarkers were the mean heart rate (MHR; [Formula: see text] = 0.04), the standard deviation of the heart rate (SD; [Formula: see text] = 0.01), and the standard deviation of NN intervals (SDNN; [Formula: see text] = 0.03) for ECG signals, and the mean breath rate (MBR; [Formula: see text] = 0.002), the standard deviation of the breath rate (SD; [Formula: see text] < 0.0001), the root mean square of successive differences (RMSSD; [Formula: see text] < 0.0001) and SDNN ([Formula: see text] < 0.0001) for RSP signals. This work extends the existing literature on the relationship between stress and HRV/BRV by being the first to introduce a transitional phase. It contributes to systematically processing mental and emotional impulse data in humans measured via ECG and RSP signals. On the basis of these identified biomarkers, artificial-intelligence or machine-learning algorithms, and rule-based classification, the automated biosignal-based psychological assessment of patients could be within reach. This creates a broad basis for detecting and evaluating psychological abnormalities in individuals upon which future psychological treatment methods could be built using portable and wearable devices.