This is a quick summary/digest on EEG (for personal future reference only). Content credits include (use these as factual references instead):

Introduction

Electroencephalography (EEG) measures voltage fluctuations resulting from Ionic currents within the neurons of the brain (Sum of groups of neurons in localized areas a opposed to individual neurons). In clinical contexts, EEG refers to the recording of the brain’s spontaneous electrical activity over a period of time – as recorded from multiple electrodes placed on the scalp.

Diagnostic applications normally focus on the frequency spectral content of EEG, to identify different types of neural oscillations (i.e. brain waves)

Applications

Used in the diagnosis of:

  • Epilepsy
  • Sleep disorders
  • Coma
  • Encephalopathies
  • Brain Death

Also used in the past in the diagnosis of:

  • Tumors
  • Stroke
  • other focal brain disorders.

But largely replaced by MRI and CT scanning technology

Mechanisms

The brain’s electrical charge is maintained by millions of neurons. The electrical potential generated by an individual neuron is far too small to be picked up by EEG. EEG Activity therefore always reflects the summation of the synchronous activity of thousands or millions of neurons that have similar spatial orientation.

If the cells do not have similar spatial orientation, their ions do not line up to create waves to be detected

Also, because voltage fields gradients fall off with the square of distance, activity from deep sources is more difficult to obtain than currents near the skull

Scalp EEG activity shows oscillations at a variety of frequencies – brain waves. Several of these oscillations have characteristic frequency ranges, spatial distributions and are associated with different states of brain functioning (e.g. walking and the various sleep stages). These oscillation or waves represent synchronized activity over a network of neurons.

EEG reflects correlated synaptic activity caused by post-synaptic potentials of cortical regions.

Methods

In scalp EEG, normally the recording is obtained by placing electrodes on the scalp with conductive gel (or paste). Usually after preparing the area with light abrasion to remove dead cells. Many systems use individual electrodes, others use caps or nets with embedded electrodes (mostly when high-density electrode arrays are needed).

Electrodes Placement

Electrode locations are largely standardized by the International 10-20 system.

Name reflects the fact that electrodes are placed at intervals that are 10% or 20% of the distance between landmarks in the skull.

  • Front to back landmarks: nasion to inion.
  • Left to right landmarks: pre-auricular points.

For higher density electrode arrays the 10-10 system is used: places electrodes at 10% intervals.

Channel Count

In most clinical applications 19 recording electrodes (plus ground and system reference) are used.

Signal’s Amplitude (Scalp EEG)

A typical adult human EEG signal is about 10 uV - 100 uV in amplitude when measured from the scalp (in optimal conditions).

However 1 uV to 100 uV is a more realistic conservative estimate under most typical cases.

EEG acquisition systems should be designed to operate with a 1 uV to 1 mV dynamic input range.

Signal’s Noise

Noise amplitude is usually 2x to 10x larger than signal amplitude.

Noise can be from multiple sources:

  • Ambient noise (external): 50/60HZ main-line hum, fluorescent lights, RF high-frequency noise from cell-phones and other RF sources (aliased when under-sampled).

  • Biopotentials noise (internal): Noise from other biopotentials e.g. EMG, EOG, ECG, EDA …etc. These are very often greater in amplitude than EEG signals.

In theory these are all common-mode signals, so should be removed by good common-mode-rejection (CMR) technicques. Thus the importance of high CMRR (common mode rejection ratio) in EEG acquisition systems

Electrodes Montages

  • Sequential Montage (also called bipolar): Each channel represents the difference between two adjacent electrodes.

  • Referential Montage: Each channel represents the difference between a certain electrode and a designated “reference” electrode. There is no standard position for this electrode but it is normally at a different position than the recording electrodes.

  • Average Reference Montage: The outputs of the amplifiers are summed and averaged and this averaged signal is used as the common reference for each channel

  • Laplacian Montage: Each channel represents the difference between an electrode and a weighted average of the surrounding electrodes.

With digital EEG, all signals are typically digitized in a particular (usually referential) montage, since any other montage can be constructed mathematically from any other.

Limitations of EEG

Several limitations:

  • Most importantly poor spatial resolution: EEG is most sensitive to a particular set of post-synaptic potentials – those generated in superficial layers of the cortex. In contrast, those generated deeper in the cortex have far less contributions to EEG.

  • EEG recordings do not directly capture axonal action potentials.

  • EEGs show a preference for activity on populations of parallel dendrites, transmitting current in the same direction and at the same time.

Thus, EEG provides information with a large bias to selected neuron types, and generally should not be used to make general claims about localized brain activity e.g. the meninges, cerebrospinal fluid and skull “smear” the EEG signal, obscuring its source.

EEG Clinical Standards/Organizations

There are three main organizations that publish standards ruling EEG acquisition systems:

  1. International Federation of Clinical Neurophysiology (IFCN): main widely cited standard though rather dated (1998).

  2. American Clinical Neurophysiology Society (ACNS): Mostly focused at clinicians and medical researchers, less focus on hardware (references IFCN).

  3. International Electrotechnical Commission (IEC): widely used among device manufacturers, only paid access. Standards retrieved from applicable data in open access publications.

EEG System’s Design solutions

This is a rough very general description of the two opposite ends of design approaches used widely in industry:

  1. The traditional discrete “mostly analog” approach: no longer viable.

    • Instrumentation Amplifiers Inputs: (High Zin, high CMRR)
    • Analog High Pass Filter (0.1Hz): reject dc electrode offset.
    • Low-noise amplifier
    • Anti-aliasing Low-pass filter (~100Hz): necessary before SAR AD conversion.
    • Sample and hold circuit (to prevent time skewing of EEG signals).
    • Analog Multiplexer.
    • 12-16 bit SAR AD conversion (single channel).

    Cons: Typically greater component count, greater PCB area, greater power consumption, greater cost and lower overall performance (component’s miss-match in path to AD conversion) as compared to newer more integrated multi-channel delta-sigma ADC solutions.

  2. The multi-channel delta-sigma ADC approach: more widely used in industry today

    Delta-sigma ADCs have become very affordable and their data rates and effective resolution have continued to increase. Higher levels of integration have made these almost a single-chip solution.

    • Differential I/O On-chip Programmable Gain Amplifiers: High Zin, Low noise.
    • Bias-drive amplifier (also called Driven Right Leg): for setting common-mode DC voltage (reducing electrode’s offset) and increasing system CMRR.
    • on-chip 16-24 bit Delta-sigma multi-channel differential ADCs: Relaxes necessity of anti-aliasing high-order low pass filters at ADC inputs, differential conversion increases CMRR.

    The low amplitude, low frequency bandwidth of EEG signals, makes delta-sigma ADCs (typically higher-resolution yet lower speed than SAR ADCs) especially well suited to this application.



Leave your comments below (or comment directly here).

Thank you for your feedback.