Human Emotion Characterization by Heart Rate Variability Analysis Guided by Respiration

Area of Science:

• Cardiology
• Psychophysiology
• Signal Processing

Background:

• Identifying emotions through physiological signals is crucial for clinical applications, particularly in diagnosing psycho-neural disorders.
• Heart rate variability (HRV) analysis is a key non-invasive technique for assessing autonomic nervous system function.
• Current HRV analysis methods may not fully capture the interplay between cardiac activity and respiratory patterns.

Purpose of the Study:

• To develop and validate a novel method for emotion identification using respiration-guided HRV analysis.
• To improve the accuracy of emotion discrimination compared to traditional HRV analysis techniques.
• To explore the potential clinical impact of this method in diagnosing psycho-neural illnesses.

Main Methods:

• A new HRV analysis method was developed, redefining the high-frequency (HF) band to be centered and have a bandwidth dependent on the respiratory spectrum.
• The method was initially tested on simulated HRV signals to evaluate estimation errors against established methods.
• The validated method was applied to a video-induced emotion database, analyzing five emotional states: relax, joy, fear, sadness, and anger.

Main Results:

• Simulated HRV analysis demonstrated minimum estimation errors, outperforming classic and HF band-centered definitions across various sympathovagal ratios.
• The proposed method successfully discriminated between different emotional states, including joy versus relax, joy versus negative emotions, and fear versus sadness (p ≤ 0.05, AUC ≥ 0.70).
• A significant correlation between HRV and respiration spectra was observed, highlighting the importance of respiratory information.

Conclusions:

• The proposed respiration-guided HRV analysis method offers improved accuracy in human emotion characterization.
• Integrating respiratory information into HRV analysis can enhance the discrimination of emotional states.
• This approach holds potential for advancing the diagnosis of psycho-neural conditions in clinical practice.