Abstract

In this work, we investigate the impact of lowpass filters on two entropy methods, Permutation Entropy (PE) and Entropy of Difference (EoD), using simulated noise signals. Colored noise, specifically white, pink and brown noise, was generated and filtered with passband frequencies of 30 Hz, 40 Hz and 50 Hz, alongside unfiltered signals. The PE and EoD values were computed to analyze the effects of filtering. The results indicate that both entropy measures decrease with lower passband frequencies. PE can effectively distinguish between pink and brown noise with and without a lowpass filter, while EoD shows similar differentiation only with filtering. These findings highlight the sensitivity of entropy measures to lowpass filtering, with implications for their application in EEG analysis.