Comment on 'Interpretation of the Lempel-Ziv Complexity Measure in the context of Biomedical Signal Analysis'

TL;DR

This paper critically examines the interpretation of the Lempel-Ziv complexity measure in biomedical signals, highlighting its dependence on harmonics, frequency, and amplitude modulation, and challenging previous assumptions.

Contribution

It provides a detailed critique of prior claims about LZ complexity's insensitivity to harmonics and frequency, offering new insights into its dependencies.

Findings

LZ complexity depends on harmonics and amplitude modulation.

LZ complexity varies with frequency, especially in moderate and high regimes.

Previous claims about LZ's insensitivity are not universally valid.

Abstract

In this Communication, we express our reservations on some aspects of the interpretation of the Lempel-Ziv Complexity measure (LZ) by Mateo et al. in "Interpretation of the Lempel-Ziv complexity measure in the context of biomedical signal analysis," IEEE Trans. Biomed. Eng., vol. 53, no. 11, pp. 2282-2288, Nov. 2006. In particular, we comment on the dependence of the LZ complexity measure on number of harmonics, frequency content and amplitude modulation. We disagree with the following statements made by Mateo et al. 1. "LZ is not sensitive to the number of harmonics in periodic signals." 2. "LZ increases as the frequency of a sinusoid increases." 3. "Amplitude modulation of a signal doesnot result in an increase in LZ." We show the dependence of LZ complexity measure on harmonics and amplitude modulation by using a modified version of the synthetic signal that has been used in the original paper. Also, the second statement is a generic statement which is not entirely true. This is true only in the low frequency regime and definitely not true in moderate and high frequency regimes.