🔗 The signal chain
Send a brain signal through the skull, noise, amplifier, filter and digitiser, one stage at a time.
Turn each step on or off and watch what it does to the signal. The steps run in this order, left to right.
- weaker, smoother
- blink, muscle, mains hum, drift
- subtracts the reference
- keeps 0.5–40 Hz
- Digitiser
What ends up in the file (2 seconds)
the brain signal at the cortex, scaled to fit
the signal after the steps
the digitiser's readings
How to read this
- Skull and scalp: the skull resists current, so the brain's signal arrives many times weaker and blurred, which takes the sharp edges off it.
- Noise: an eye blink (a big slow bump), a burst of jaw muscle (fast fuzz), mains hum at 50 Hz and slow drift from the electrodes all add on top.
- Differential amplifier: each channel is electrode minus reference. Mains hum reaches both nearly equally, so most of it cancels. A blink is bigger at the front, so it does not cancel.
- Filter: keeps 0.5–40 Hz. Drift, muscle fuzz and what is left of the hum mostly go; the blink stays, because it is slow like brain waves.
- Digitiser: takes readings at a fixed rate. Below twice the highest frequency left in the signal, fast waves fold back as false slow waves (aliasing). Try 64 a second with the filter off.
Key terms
- Volume conduction
- Current spreading through tissue and bone on its way to the scalp.
- Common-mode noise
- Noise that reaches the electrode and the reference equally, such as mains hum.
- Band-pass filter
- A filter that keeps a range of frequencies and drops the rest.
- Sampling rate
- Readings per second (Hz).
- Nyquist rule
- Sample at more than twice the highest frequency in the signal, or it aliases.
Sources
- Nunez P. L., Srinivasan R. (2006). Electric Fields of the Brain: The Neurophysics of EEG, 2nd ed. Oxford University Press.
- Sinha S. R. et al. (2016). American Clinical Neurophysiology Society Guideline 1: minimum technical requirements for performing clinical electroencephalography. J Clin Neurophysiol 33(4): 303–307.
- Urigüen J. A., Garcia-Zapirain B. (2015). EEG artifact removal: state-of-the-art and guidelines. Journal of Neural Engineering 12(3): 031001.
Used in lessons 3, 8, 9 of EEG from zero.