Digital Polysomnography

Frank Walther

LEARNING OBJECTIVES

On completion of this chapter, the reader should be able to:

1. List the sources of all signals.
2. Describe how differential amplification effects signal voltage and polarity.
3. Discuss the source localization of electroencephalographic signals.
4. Apply Ohm’s law and Nyquist’s sampling theory to digital polysomnography (PSG).
5. Show the order in which signals are processed in PSG systems.
6. Discuss the filtering of PSG signals.
7. Outline the concepts of calibration.

The word polysomnography in its root derivation means many (poly) sleep (somno) writings (graphy). A polysomnogram is produced by a multiple-channel recording instrument comprised of a hardware device (often called an “amplifier”) that interfaces with software to produce a digital record of multiple biophysical variables during the course of a sleep session.

Historically, polysomnographic (PSG) recordings were conducted with analog equipment that traced this physiologic activity by mechanical means to paper. Although these paper-based systems performed admirably, the machines were large and resource intensive. They required replenishment of paper and ink and ongoing maintenance of all the mechanical parts, not to mention extraordinary storage requirements. Because the activity was converted by mechanical means, each study required precise mechanical baseline calibrations. Scoring required page-by-page examination of each epoch, and event marking was done by hand. Present-day digital systems afford sleep professionals all the benefits of software programming and networking. Studies can be viewed on multiple screens, allowing for simultaneous acquisition and scoring. Montages can be adjusted across all parameters during and after the recording, video can be collected and synchronized with the recording, and indices can be estimated or tabulated in real time. Physicians are able to view studies remotely in real time and in review mode. Reports can be tailored to suit the needs of the health care provider.

The quality and validity of signal processing remains intact in modern digital PSG as long as adequate PSG system specifications and good PSG skills are observed. The American Academy of Sleep Medicine (AASM) Scoring Manual (1) has defined standards and best practices (i.e., sampling rates, bit rates, screen resolution, etc.) for the recording process to ensure that all studies accurately reflect the patient condition (see Table 34-1).

Table 34-1 Digital Specifications (1)

Maximum EEG and EOG electrode impedance	5 kΩ
Minimum digital resolution	12 bits per sample
Digital screen resolution	1,600 × 1,200
Notch filters 50/60 Hz	Per channel
Time scale range window view	5 s to entire recording
EEG, electroencephalography; EOG, electrooculogram.

PRINCIPLES OF ELECTRICAL CONDUCTION

Signal Sources

The basic function of a PSG system is to record signals associated with specific physiologic parameters and convert this activity into visible tracings that can be measured and analyzed. In PSG, there are three sources for these signals:

Bioelectric potentials
Transduced signals from sensors attached to the patient
Signals derived from ancillary equipment

Bioelectric potentials are voltages generated by living tissue. Examples of bioelectric potential recordings include the electroencephalogram (EEG), electrooculogram, electromyogram, and electrocardiogram (ECG). Bioelectric signals are recorded using surface electrodes attached directly to the patient’s skin over the area of interest. For the purposes of a discussion regarding the general pathway of these source signals, we concentrate on EEG recordings of the brain.

The brain’s electrical charge is the result or summation of the electrical activity of billions of neurons. Neurons are electrically charged (or polarized) by transport proteins that pump ions across their cell membranes. When a neuron receives a signal from its neighbor, an action potential is triggered; it responds by releasing ions into the synaptic space outside the cell. Ions of like charge repel each other, and when many ions are pushed out of many neurons at the same time, they can push their neighbors, who push their neighbors, and so on, in a wave. When the wave of ions reaches the electrodes on the scalp, they can push electrons through the measuring circuit comprising the electrodes and a sensitive amplifier. Thus, the measuring circuit can measure the difference in this electron “push,” or voltage, between any two electrodes. Recording these voltage differences over time gives us the EEG (2). Note that the EEG does not directly measure electrical impulses in the cells, but instead the EEG arises from changing ionic concentrations and the resulting charge in the extracellular space.

The electric potentials generated by a single neuron are far too small to be picked by the EEG (3), because the skull and skin act as capacitors and dampen the signal. EEG activity, therefore, always reflects the summation of the synchronous activity of thousands or millions of neurons that have similar spatial orientation (4).

Transduced signals are voltages usually supplied as DC voltages by sensors that are attached to the patient as opposed to voltages that are produced by the body. This can include position sensors, snore sensors, and respiratory sensors.

Ancillary equipment, such as continuous positive airway pressure devices, pulse oximeters, and end-tidal CO₂ (EtCO₂) monitors, can be interfaced with the PSG system by using one of many available DC inputs. The industry trend is to integrate pulse oximetry with the amplifier unit.

Electrodes

The first step of the signal pathway external to the body is the electrode. The electrodes, secured using a conductive paste, are the conduits for electron push and, as such, must have, as one of their properties, high electric conductivity. As seen in Table 34-2, the conductivities of copper, silver, and gold are all comparable.

These elements are all from group 12 of the periodic chart and are heavy, dense, malleable metals with high electric conductivity. This high conductivity is due to an atomic makeup of having free outer shell electrons in relation to a dense nucleus. The electrode cups commonly used in PSG are typically silver or gold plated. Those two metals have the additional desirable property of being resistant to the corrosive effects of oxygen, H₂O, and bodily fluids such as sweat. Gold is seen as superior because it optimizes contact impedance over the long duration of a sleep study. Silver-silver chloride and disposable electrodes made of these metals or plastic can also be used.

SIGNAL PATHWAY TO THE DIGITAL PSG

The signals previously discussed, whether they originate from the patient or are transduced or ascribed to the patient, follow a pathway that we shall refer to as a “channel.” These input signals are amplified and processed, so that they are strong enough and clean enough to be converted into a digital signal that is ready for visual display, storage, and analysis.

Bioelectric signals are amplified with differential amplifiers, which reduce recording artifacts by subtracting
voltages that are common to an electrode pair. The electrodes are usually routed to the amplifier through a small portable junction box, sometimes referred to as the “jack-box” or the “headbox.” The industry trend is to integrate these functions even further by placing the jackbox, amplifier, and digitizer all in the same enclosure. A differential amplifier schematic is shown in Figure 34-1.

Table 34-2 Conductivity

Metal	Conductivity σ(1/Ωm) at 20° C
Silver	6.29 × 10⁷
Copper	5.95 × 10⁷
Gold	4.52 × 10⁷

Figure 34-1 Differential amplifier.

OHM’S LAW

The three electric properties that are the variables in the equation known as Ohm’s law are voltage, current, and resistance. Ohm’s law is commonly represented as

V = IR or voltage = current × resistance.

Ohm’s law can be rearranged to show current as a function of voltage and resistance according to the formula I = V/R, and from this formula, one can deduce that under conditions of constant voltage, if resistance is increased, current or flow of electrons will be decreased. This concept has practical implications when we consider the fact that in our signal pathway, there are several factors that contribute to higher resistance. This means that the technologist must be vigilant in reducing these effects. Examples of things that could contribute to higher resistance are electrode cups inadequately filled with conductive paste, electrode damage, electrode wire damage, poor patient skin preparation, and incomplete removal of lotions and other skin products.

Notice that what is being measured is the difference in voltage of the two input electrodes as determined by their relation to the reference electrode (Fig. 34-1). The first input electrode is the exploring electrode. In PSG, we commonly refer to the mastoid electrodes as the reference electrodes because we have placed them on distant bony electrically neutral sites (M1 and M2). The M1 and M2 electrodes are considered the second input electrodes. The referential electrode is labeled “Vref.” Common ground is an additional connection (usually separate from the reference) and is necessary for equalizing electric potentials of the patient and the input circuits of the amplifier. Also, notice the power supply voltage, represented in the diagram by Vs in and Vs out. The power supply is necessary to carry out the function of amplifying the faint biopotential signals.

An essential function of the differential amplifier is that of common mode rejection, or in plain words, the cancelation of unwanted voltages that are common to both input electrodes. The ability of the differential amplifier to perform this vital function is expressed as the common mode rejection ratio (CMRR). This is the ratio of the differential voltage gain to the common mode voltage gain. If the differential amplifier were perfectly symmetrical, the common mode gain would be zero because both voltages would effectively cancel each other out, and the CMRR would be infinite. However, in reality, there is a slight amount of common voltage gain, resulting in industry standard CMRRs of 10,000 to 1 or even as high as 100,000 to 1.

SIGNAL PROCESSING

Signal Polarity and Summation

The processing of bioelectric signals includes conversion from analog to digital signal format. Electrodes on the surface of the scalp measure signals of constantly changing voltage, or amplitude, as a function of time. Differential amplifiers process the signals from two electrodes, an exploring electrode and a reference. Because the electrodes are positioned at various locations (according to the 10-20 system), the resulting tracing will vary according to the position of the electrodes and their distance and proximity to the source of voltage (see Fig. 34-2).

In EEG, it is conventional to represent negative voltage from a single input in relation to a reference as an upward deflection and, vice versa, a positive voltage is represented as a downward deflection. This relationship
of the tracing to the electric zero above or below the baseline is referred to as “polarity,” as illustrated in Figure 34-3.

Figure 34-2 Differential amplification of exploring and reference electrodes.

Figure 34-3 Polarity.

With differential amplifiers, signal tracings shown on screen are the result of subtracting the voltage contributions from two input electrodes. This is accomplished by inverting the signal from the reference electrode. The formula for combining the signals is exploring electrode (V1) voltage minus reference electrode (V2) voltage (V1-V2), which we refer to as voltage drop.

In summation, under ideal conditions, the variables that determine bioelectric tracings are as follows:

Polarity
Voltage
Source
Electrode proximity
Time

Please examine the simple case of two electrodes detecting equal polarity and voltage (50 µV) at the same point in time, at the same distance from the source, in Figure 34-4. We apply the formula for voltage drop 50 – (50) = 0 and see by visual examination the result is no signal.

Now follow along with the case of simultaneous charges of equal polarity and equal proximity but differing voltages. 50 – (25) = 25. In this case, we see the resulting tracing having the same polarity (Fig. 34-5), but if the reference electrode is higher in voltage than the input voltage, then the polarity flips. Now take the case of simultaneous charges of opposite polarity, equal proximity, and same voltages: 50 – (-50) = 100, as illustrated in Figure 34-6.

Figure 34-4 Two electrodes with equal polarity and voltage input.

Figure 34-5 Two electrodes with equal polarity and different voltage inputs.

The number of possible combinations is infinite, but we will discuss one more example. That is the combination of two similar polarities, similar signal amplitudes but making their contributions at differing times. By using our formula, V1 – V2 as we go along the axis of time, we see the resulting (and recognizable) signal in Figure 34-7.