A Non-contact Framework for Shortness of Breath Recognition and effects on Health Parameters during HCI

Breath is vital for survival and plays an essential role in enhancing the physical, mental, and spiritual well-being of a human. Real-time breath pattern monitoring during Human-Computer Interaction (HCI) help in the diagnosis and potential avoidance of various health problems. However, state-of-the-art approaches for breath monitoring are usually contacted basis and/or limited to medical facilities. In this paper, the goal of the study is to investigate the non-contact measurement technique of breath pattern, and a framework has proposed for recognizing the shortness of breath, conscious breath, and deep breath wave patterns while watching various emotional video stimuli. The proposed framework consists of an FMCW Doppler radar sensor and an analytical algorithm to calculate the tidal volume (TV), which is proportional to chest displacement. We considered the TV as the threshold amplitude of both inhalation and exhalation for various breath wave patterns. Based on the threshold value, we categorized the breath wave patterns into three different classes such as shortness of breathing, conscious breathing, and deep breathing for both inhalation and exhalation. We also measured some health signatures such as abdomen, airflow, ECG, EMG, EEG, heart rate, and SpO2 concerning various breath waveform during HCI. For validating our radar breath data, we used one 32-bit PSG channel device on contact mode. The obtained PSG data is of a higher sampling rate (256Hz) compared to that of the radar sampling rate(40Hz). Hence, we downsampled the PSG data. We used the Pearson correlation coefficient between the chest displacement of radar data and the thoracic displacement of PSG data. The obtained experimental results demonstrate the effectiveness of our proposed approach.