Robust algorithmic detection of the developed cardiac pathologies and emerging or transient abnormalities from short periods of RR data

Numerous research efforts and clinical testing have confirmed validity of heart rate variability (HRV) analysis as one of the cardiac diagnostics modalities. The majority of HRV analysis tools currently used in practice are based on linear indicators. Methods from nonlinear dynamics (NLD) provide more natural modeling framework for adaptive biological systems with multiple feedback loops. Compared to linear indicators, many NLD‐based measures are much less sensitive to data artifacts and non‐stationarity. However, majority of NLD measures require long time series for stable calculation. Similar restrictions also apply for linear indicators. Such requirements could drastically limit practical usability of HRV analysis in many applications, including express diagnostics, early indication of subtle directional changes during personalization of medical treatment, and robust detection of emerging or transient abnormalities. Recently we have illustrated that these challenges could be overcome by using classific...