arXiv:2604.13815v3 Announce Type: replace Abstract: Heart rate variability (HRV) analysis is important for the assessment of autonomic cardiovascular regulation. The inverse Gaussian process (IGP) has been widely used for beat-to-beat HRV modeling. However, dynamic estimation of the IGP parameters remains a critical challenge. Conventional parametric approaches such as autoregressive models might be insufficient to capture the nonlinear characteristics of R-R series. In this study, we investigated whether recurrent neural networks (RNNs) can be used for IGP parameter identification and thereby enhance probabilistic modeling of R-R dynamics. Four representative RNN architectures, namely, GRU, LSTM, Structured State Space sequence model (S4), and Mamba, were compared and evaluated using the Kolmogorov-Smirnov statistics. The results demonstrate the possibility of combining RNN with the IGP for beat-wise R-R series modeling. This approach provides a flexible basis for probabilistic HRV modeling and for future incorporation of more complex physiological mechanisms and dynamic conditions.