A Pilot Study of Wearable Respiratory Driven Right and Left Nostril Breathing
Abstract
Analysis of breathing patterns reveals factors influencing nasal airflow passages. Shiva Swarodaya in Swara Yoga texts emphasizes the importance of right and left nostril breathing practice. A non-invasive and wearable nasal mask that continuously monitors the breathing parameters using temperature, pressure, and humidity sensors, providing real-time data from both nostrils is developed. The study focuses on the distinction of airflow signals from two nostrils and classifies the individuals as active-right and active-left nostril breathing by leveraging artificial intelligence (AI)-driven machine learning classification models for the classification and prediction. The statistical features, such as, mean, median, standard deviation, skewness, and kurtosis along with Teager energy and Shannon’s entropy calculated from every 30 sec segments of temperature, pressure, and humidity signals collected from both nostrils were used for classifying the breathing as active-right-nostril-breathing and active-left-nostril-breathing. k-nearest neighbors (KNN) algorithm, random forest (RF), and support vector machine (SVM, linear, polynomial, Radial Basis Function were evaluated using accuracy, precision, recall, and F1-score. SVM-RBF achieved the highest accuracy of 98.06, precision of 0.98, recall of 0.98 and F1-score of 0.98. Accurate nostril dominance classification aids in optimizing performance, nervous system function, and stress management. The developed algorithm revealed that right-nostril breathing increases breathing rate, tidal volume, and minute ventilation, with shorter inhalation and exhalation time. The web-app is developed for respiratory insights and vital feature predictions.