Application of nonlinear methods to heart rate variability signal analysis

Correlation dimension (D/sub 2/) with spatial embedding and Poincare plots are employed in the nonlinear analysis of heart rate variability among 8 normal subjects, 9 rheumatoid cardiopathy patients, and 6 coronary artery disease patients. Rheumatoid cardiopathy patients and coronary artery disease patients have lower D/sub 2/ compared to the normal subjects. This can serve as a discriminative measure. A proper epoch (N) of the spatial embedding method is 20,000 R-R intervals in this study, upon which the D/sub 2/ value of normal subjects is 5.16. In the analysis of Poincare plots, the normal subjects totally exhibit longer long-axel (L), smaller included angle (/spl alpha/), and relatively regular shape. The bradycardia probability of rheumatoid cardiopathy patients is apparently bigger than the normal's. The coronary artery disease patients show a relatively bigger value of longer long-axel (L). The results and conclusions have been applied in our Holter Analysis System.