A new methodology for nonlinear heart function analysis: Studying just the beat morphology

Nonlinear analysis techniques have been used to study the heart rate variability series (which only contains RR interval information) and the ECG (which contains both RR interval and beat morphology information). It is still an open issue whether the RR series is chaotic or stochastic. Nonlinear properties of ECG are less controversial, despite containing heart rate information. In this work, we propose a new methodology to achieve phase space reconstruction that only uses beat morphology information; thus it avoids possible phase space contamination from stochastic components of the RR intervals. Nonlinear statistics computed with this approach are able to discriminate between healthy and congestive heart failure patients.