Argentina Leite

Abstract
This work focus on detection of diseases from Heart Rate Variability (HRV) series using Long Short-Term Memory (LSTM) networks. First, non-linear models are used to extract sequences of features that characterize the HRV series. These time sequences are then used as input for the LSTM. HRV recordings from the Noltisalis database are used for training and testing this approach. The results indicate that the procedure provides accuracy scores in the range of 86.7% to 90.0 % on the test set. © 2020 IEEE.

Abstract
The concept of health indicates physical, mental and social well-being. Psychological stress is commonly present among freshmen due to social and environmental changes. An approach to study the impact of stress on students relied on biological data assessment. In this work, electrocardiogram signals from first year students, from the Biomedical Engineering course, were collected during an oral presentation, acquiring the RR time series. Linear and nonlinear methodologies are used to extract features that best characterize the RR time series in young subjects under stress conditions.

Abstract
Several Heart Rate Variability (HRV) based novel methodologies for describing heart rate dynamics have been proposed in the literature with the aim of risk assessment. One such methodology is ARFIMA-EGARCH modeling which allows the quantification of long range dependence and time-varying volatility with the aim of describing non-linear and complex characteristics of HRV. This study applies the ARFIMA-EGARCH modeling of HRV recordings from 30 patients of the Noltisalis database to investigate the discrimination power of a set of features comprising currently used linear HRV features (low and high frequency components) and new measures obtained from the modeling such as, long memory in the mean, and persistence and asymmetry in volatility. A subset of the multidimensional HRV features is selected in a two-step procedure using Principal Components Analysis (PCA). Additionally, supervised classification by quadratic discriminant analysis achieves 93.3% of discrimination accuracy between the groups using the new feature set created by PCA. © 2018 IEEE.

Abstract
Heart Rate Variability (HRV) data exhibit long memory and time-varying conditional variance (volatility). These characteristics are well captured using Fractionally Integrated AutoRegressive Moving Average (ARFIMA) models with Generalised AutoRegressive Conditional Heteroscedastic (GARCH) errors, which are an extension of the AR models usual in the analysis of HRV. GARCHmod-els assume that volatility depends only on the magnitude of the shocks and not on their sign, meaning that positive and negative shocks have a symmetric effect on volatility. However, HRV recordings indicate further dependence of volatility on the lagged shocks. This work considers Exponential GARCH (EGARCH) models which assume that positive and negative shocks have an asymmetric effect (leverage effect) on the volatility, thus better copping with complex characteristics of HRV. ARFIMA-EGARCH models, combined with adaptive segmentation, are applied to 24 h HRV recordings of 30 subjects from the Noltisalis database: 10 healthy, 10 patients suffering from congestive heart failure and 10 heart transplanted patients. Overall, the results for the leverage parameter indicate that volatility responds asymmetrically to values of HRV under and over the mean. Moreover, decreased leverage parameter values for sick subjects, suggest that these models allow to discriminate between the different groups. © 2016 CCAL.

Abstract