Statistical and Machine Learning Approach to Study the Smoking Impact on Heart Rate Variability Features

• Published in: 2021 International Conference on Emerging Smart Computing and Informatics (ESCI) pp. 170 - 174
• Main Authors: Rathod, SR, Chaskar, UM, Patil, CY
• Format: Conference Proceeding
• Language: English
• Published: IEEE 05.03.2021
• Subjects: Autonomous Nervous System(ANS); Classification algorithms; Electrocardiogram(ECG); Electrocardiography; Feature extraction; Heart rate variability; Heart Rate Variability (HRV); Machine Learning(ML); Modulation; Prediction algorithms; Statistical analysis; Synthetic Minority Oversampling Technique(SMOTE)
• DOI: 10.1109/ESCI50559.2021.9396936

In the present study, we studied the impact of cigarette smoking on Heart Rate Variability(HRV). HRV is a non-invasive screening technique that measures the variation in consecutive heartbeats. HRV features were derived from the Electrocardiogram(ECG), and it is an indicator of the autonomous nervous system(ANS) status. The ECG was recorded before and 5 minutes after smoking, and HRV features were extracted. The healthy subjects HRV data was considered as a baseline for the study. Both real and synthetic HRV data were used to perform this study. The synthetic HRV data was generated using Synthetic Minority Oversampling Technique(SMOTE) Machine Learning(ML) technique. The data of healthy and smokers was analyzed using both statistical test and machine learning algorithms. Both approaches suggest an ANS imbalance in smokers that ultimately shows the modulation in HRV. The modulation in HRV indicates the high chances of cardiovascular risk in smokers.