Neuro-fuzzy modelling of heart rate signals and applications to diagnostics

Abstract: Heart rate variability (HRV) is examined by methods combining neural networks and fuzzy logic. Multiple features are extracted from examples of heart rate data from normal adult subjects, subjects recently suffering from a heart attack, subjects with a history of ischaemic heart disease undergoing a coronary investigation, and subjects in atrial fibrillation. Special attention is given to the analysis of fractal features extracted from heart rate sequences. The methodologies of fuzzy neural networks (FuNN) and evolving fuzzy neural networks (EFuNN) are described and applied to heart rate variability. A description of applications of heart rate variability analysis in medicine is given. The proposed methods can be used for further practical applications in this area. Key words: heart rate variability, neural networks, fuzzy memberships, evolving networks, fractal measures.