STATISTICS IN MEDICINE Statist. Med. 2002; 21:2225–2242 (DOI: 10.1002/sim.979) Potential of feature selection methods in heart rate variability analysis for the classication of dierent cardiovascular diseases Agnes Schumann 1 , Niels Wessel 1 , Alexander Schirdewan 2 , Karl Josef Osterziel 2 and Andreas Voss 1;∗;† 1 University of Applied Sciences; Jena; Germany 2 Franz-Volhard-Hospital; Humboldt University; Berlin; Germany SUMMARY In this study heart rate variability (HRV) analysis was applied to characterize patients suering from coronary heart disease (CHD), dilated cardiomyopathy (DCM) and patients who had survived an acute myocardial infarction (MI). On the basis of several HRV parameters, an optimal discrimination between the dierent kinds of cardiovascular diseases and between the diseases and healthy controls (HC) was derived by feature selection and linear classication. For each task a small favourable subset of a set of 33 potentially interesting HRV measures was selected with the intention of improving the diagnostic value and facilitating the physiological interpretation of HRV analysis. Time- and frequency-domain parameters as well as parameters from non-linear dynamics were included in the analysis. With the expectation that dierent diseases are characterized by dierent phenomena, feature selection was applied for each task separately. Using the features optimal for one task to another task can reveal a loss in performance, but it turned out that one specic parameter set (set1: normalized low frequency LF= P and a non-linear variability measure WPSUM13) was applicable for all tasks, where diseased and healthy subjects have to be distinguished, without signicant reduction in performance. This set seems to be a general marker for pathologic changes in HRV and might be used for early detection of heart diseases. The classication between dierent heart diseases requires another parameter set (set2: meanNN and sdaNN, reecting the steady state behaviour of the heart rate and long and short term SEAR describing the spectral composition). However, the use of set1 for the separation of dierent kinds of diseases, where set2 is appropriate, led to signicant reduction in performance and vice versa. This observation may be important for future developments of HRV measures especially suitable for the assessment of disease severity. Copyright ? 2002 John Wiley & Sons, Ltd. 1. INTRODUCTION The analysis of heart rate variability (HRV) became a standard method for studying the role of the autonomic nervous system in heart control. From the analysis of amount and structure ∗ Correspondence to: Andreas Voss, Fachhochschule Jena, Carl-Zeiss-Promenade 2, 07745 Jena, Germany † E-mail: voss@fh-jena.de Contract=grant sponsor: Deutsche Forschungsgemeinschaft DFG; contract=grant number: vo505=2-3 Received March 2000 Copyright ? 2002 John Wiley & Sons, Ltd. Accepted March 2001