Use of Discrete Wavelet Transform to Assess Impedance Fluctuations Obtained from Cellular Micromotion

Abstract

Electric cell–substrate impedance sensing (ECIS) is an attractive method for monitoring cell behaviors in tissue culture in real time. The time series impedance fluctuations of the cell-covered electrodes measured by ECIS are the phenomena accompanying cellular micromotion as cells continually rearrange their cell–cell and cell–substrate adhesion sites. Accurate assessment of these fluctuations to extract useful information from raw data is important for both scientific and practical purposes. In this study, we apply discrete wavelet transform (DWT) to analyze the concentration-dependent effect of cytochalasin B on human umbilical vein endothelial cells (HUVECs). The sampling rate of the impedance time series is 1 Hz and each data set consists of 2048 points. Our results demonstrate that, in the Daubechies (db) wavelet family, db1 is the optimal mother wavelet function for DWT-based analysis to assess the effect of cytochalasin B on HUVEC micromotion. By calculating the energy, standard deviation, variance, and signal magnitude area of DWT detail coefficients at level 1, we are able to significantly distinguish cytotoxic concentrations of cytochalasin B as low as 0.1 μM, and in a concentration-dependent manner. Furthermore, DWT-based analysis indicates the possibility to decrease the sampling rate of the micromotion measurement from 1 Hz to 1/16 Hz without decreasing the discerning power. The statistical measures of DWT detail coefficients are effective methods for determining both the sampling rate and the number of individual samples for ECIS-based micromotion assays.

Keywords

• ECIS
• discrete wavelet transform
• cytochalasin B
• micromotion
• variance
• signal magnitude area