3.2 Physiologic Signals & Calibrations

The Diagnostic PSG Signal Array

A diagnostic polysomnogram (PSG) is a multi-parameter recording. The RPSGT exam expects you to know what each channel measures, which sensor scores which event, and how to confirm the signal is valid. The full AASM recommended array is summarized below.

The EOG channels are intentionally placed off-axis: E1 is 1 cm below the left outer canthus and E2 is 1 cm above the right outer canthus, both referenced to a mastoid. This offset makes conjugate eye movements appear out of phase between the two channels — the signature used to confirm rapid eye movements in REM and to distinguish true eye movement from EEG artifact.

Airflow: Thermistor vs. Nasal Pressure

This distinction is heavily tested:

• Oronasal thermistor detects temperature change between inhaled and exhaled air. It captures oral and nasal airflow and is the AASM sensor for scoring apneas (a ≥90% drop in the thermistor signal for ≥10 seconds).
• Nasal pressure transducer measures pressure swings and is far more sensitive to subtle airflow limitation. It is the AASM sensor for scoring hypopneas (a ≥30% drop for ≥10 seconds with an associated ≥3% desaturation or arousal) and for detecting flow limitation and the flattened plateau that signals upper-airway resistance.

If the thermistor fails, the nasal pressure signal (or RIP sum) is an acceptable backup for apnea scoring, and vice versa — a substitution the technologist must log.

Calibrations

Calibrations prove the equipment and sensors work before you commit a night of recording. There are two distinct kinds, and the exam tests the difference.

1. Electrical (Square-Wave) Calibration

An equal calibration voltage (for example, a 50 µV square wave) is applied to every channel before sensors are connected. Each amplifier should produce an identical, well-formed waveform with matching amplitude and the same rise and decay (time constant). Differences reveal a faulty amplifier, a wrong gain, or a misset filter. This is a test of the recording system, not the patient, and it confirms gains and filters are uniform across channels so that a 50 µV deflection means the same thing on every trace.

2. Physiologic (Bio-) Calibration

With all sensors attached and the patient awake, the technologist directs specific maneuvers and watches the screen to confirm each channel responds correctly:

The bio-calibration confirms each sensor is attached, correctly referenced, and reading from the intended source. Document it in the technologist log. Any channel that does not respond — a flat EOG on lateral gaze, a silent chin EMG on teeth-gritting — must be corrected before lights out, because these are the very channels REM and arousal scoring depend on.

Filters & Sensitivities

Filters shape which frequencies are displayed; sensitivity (gain) scales amplitude. The AASM publishes recommended settings so studies are comparable between labs.

Note the logic: the EMG LFF is set high (10 Hz) to roll off slow movement and sweat drift while preserving fast muscle bursts, whereas the EEG LFF is set low (0.3 Hz) so slow-wave (delta) activity survives. The airflow/effort HFF is low (15 Hz) because respiration is a slow signal and higher frequencies are only noise.

A 60 Hz notch filter (mains/line filter) should be the last resort for electrical noise — it can mask a real impedance or grounding problem and distorts nearby frequencies. Always correct the cause (re-prep electrode, secure the patient ground, remove an interfering device) before applying a notch filter.

Sensitivity

Typical sensitivities: EEG roughly 7 microvolts per millimeter (µV/mm), EOG ~7 µV/mm, chin EMG ~5 µV/mm or finer, ECG much larger (~1 mV scale). A key rule the exam tests: sensitivity changes only the displayed amplitude, not the underlying digitized data. Lowering EEG sensitivity to make 60 Hz artifact "look smaller" hides the problem without fixing it and can flatten a genuine slow-wave or K-complex you needed to score.

Common Trap

Applying the EMG filter (LFF 10 Hz) to an EEG channel attenuates delta and theta activity — the study will under-stage N3 and may misread true slow-wave sleep as lighter stages. Match the filter table to the channel, every time.