Approximate Entropy Based on Isometric Tremor: Reliability and Optimal Sampling Frequency

INTRODUCTION Approximate entropy (ApEn) [1] calculated from isometric tremor torque time-series provides additional information of the neuromuscular system, compared to that obtained from the traditional analysis (standard deviation (SD) or spectral analysis) [2]. The ApEn value is determined by the complex pattern of the neuromuscular system, decreases with pathology as e.g. Parkinson’s disease, and reflects the severity of the disease [2]. However, ApEn can be biased due to both overand under sampling, when calculated from continuous biological signals as e.g. isometric tremor. Methodological issues such as optimal sampling frequency and reliability of the method has not previously been assessed.