4: Filtering the EEG and ERPs

Last updated
Save as PDF

Page ID: 87945

\( \newcommand{\vecs}[1]{\overset { \scriptstyle \rightharpoonup} {\mathbf{#1}} } \) \( \newcommand{\vecd}[1]{\overset{-\!-\!\rightharpoonup}{\vphantom{a}\smash {#1}}} \)\(\newcommand{\id}{\mathrm{id}}\) \( \newcommand{\Span}{\mathrm{span}}\) \( \newcommand{\kernel}{\mathrm{null}\,}\) \( \newcommand{\range}{\mathrm{range}\,}\) \( \newcommand{\RealPart}{\mathrm{Re}}\) \( \newcommand{\ImaginaryPart}{\mathrm{Im}}\) \( \newcommand{\Argument}{\mathrm{Arg}}\) \( \newcommand{\norm}[1]{\| #1 \|}\) \( \newcommand{\inner}[2]{\langle #1, #2 \rangle}\) \( \newcommand{\Span}{\mathrm{span}}\) \(\newcommand{\id}{\mathrm{id}}\) \( \newcommand{\Span}{\mathrm{span}}\) \( \newcommand{\kernel}{\mathrm{null}\,}\) \( \newcommand{\range}{\mathrm{range}\,}\) \( \newcommand{\RealPart}{\mathrm{Re}}\) \( \newcommand{\ImaginaryPart}{\mathrm{Im}}\) \( \newcommand{\Argument}{\mathrm{Arg}}\) \( \newcommand{\norm}[1]{\| #1 \|}\) \( \newcommand{\inner}[2]{\langle #1, #2 \rangle}\) \( \newcommand{\Span}{\mathrm{span}}\)\(\newcommand{\AA}{\unicode[.8,0]{x212B}}\)

Learning Objectives

In this chapter, you will learn to:

Compare the frequency content of an ERP waveform with the frequency response function of a filter to predict how well the filter will attenuate the noise in the data
Determine the impulse response function of a filter and conceptualize filtering as a process that replaces each point in the unfiltered waveform with a scaled copy of this function
Think of an ERP waveform as a series of impulses, one at each time point
Predict how a filter will distort an ERP waveform on the basis of the filter’s impulse response function
Select filter parameters that provide the best balance between noise reduction and distortion of the waveform
Create artificial waveforms and filter them to see how a filter might be distorting your data

You must use filters in ERP experiments. At a minimum, your amplifier includes an antialiasing filter that must be used prior to digitizing the EEG. In almost all ERP experiments, additional filtering is important for reducing sources of noise that would otherwise create large measurement error and reduce your statistical power. However, when filters are misused, they can dramatically distort your data, leading to incorrect conclusions. As a result, it’s vitally important that you understand how filters work and the conditions under which they can produce significant distortion of your ERP waveforms.

For most ERP researchers, there is no topic more boring than filtering. At the core of filtering is a mathematical operation called convolution. Even the word “convolution” sounds complicated and boring!

However, you can get a reasonable understanding of filtering by seeing how convolution works visually, without ever seeing an equation. This chapter takes you through a set of exercises that will show you how convolutions are used for filtering without any equations. If you want a more detailed description of filtering, you should read Chapter 7 in Luck (2014). If you want to understand the math, you can read Chapter 12 in Luck (2014), which is available for free online.