Recurrence Quantification Analysis of RR Interval Signals of Female Smokers and Non-smokers during Different Phases of Menstrual Cycle

In this paper, Recurrence Quantification Analysis (RQA) method has been implemented on RR interval signals to study the changes in the cardiac autonomic regulation among female smokers and non-smokers across different menstrual phases. The 6 min ECG signals of 24 different female volunteers, among which 12 are smokers and 12 are non-smokers, were recorded during different stages of menstruation. The RR interval signals were extracted from these ECG Signals and the analysis of these RR intervals were performed using recurrence plot and RQA. The statistically significant RQA features extracted during each menstrual cycle were identified using a series of statistical tests (t-test, Classification And Regression Tree (CART), Boosted Tree (BT) and Random Forest (RF)). The statistically significant features were applied as input to Multilayer Perceptron (MLP)-based artificial neural networks (ANNs) to classify smokers and nonsmokers. The classifiers could provide the classification accuracies of ≥ 80% during all the different stages of menstruation, supporting the variation in the cardiac autonomic regulation among female smokers and non-smokers.