Advanced Monitoring · processed EEG
Run the processed-EEG module.
How to enable, configure, present and troubleshoot the depth-of-anaesthesia surface, using the controls exactly as they appear in the app.
Before you start
- Use an up-to-date browser on a reasonably capable phone, tablet or laptop; the surface draws a live trace and reads the signal core continuously.
- Where possible, present on a wide screen — the trace, the index and the numeric panel are easiest to read side by side.
- The module is deliberately inactive until you enable it. A fresh session shows the standard monitor only.
Enable processed EEG
- Open the EEG tab on the controller. Its panel is headed Depth of anaesthesia · processed EEG.
- Enable the module. Tap Enable processed EEG — or apply any teaching preset, which enables it and sets a coherent picture in the same tap.
- Open the surface on the monitor. The learner switches their monitor to the depth surface; until the module is enabled it shows .
Looking is not enabling: the room is untouched until you tap Enable processed EEG or apply a preset. There is no separate off switch — ending the session is how the room returns to the standard monitor.
Linked and manual modes
- Linked The default, coherent teaching mode: one latent depth drives the index, stage and suppression together, and the device modifiers bias what is displayed.
- Manual Intentionally decouples the outputs so you can build a deliberate teaching contradiction. The controller shows an unmistakable amber banner while it is on — quoted in full below.
A manual picture is a constructed teaching case. Identify it as such, and never describe it as the patient's physiology. The controller says so itself, for as long as the mode is on:
Your depth setting shapes how the trace bursts and suppresses; the number is then counted back from what the trace actually did over the last 30 seconds. That is why it trails a depth change rather than snapping to it — teach the lag rather than waiting it out. It is a teaching value, not a validated measurement.
Presets and transitions
One tap configures the whole picture. The controller offers:
- Adequate GA An on-target adult theatre picture.
- Too light An awareness-risk picture; check delivery and context.
- Profound depth Excess hypnotic effect.
- Burst suppression Recognise suppression.
- Isoelectric Profound suppression with good contact.
- EMG / NMBA A two-phase confound: high EMG falsely elevates the index, and giving an NMBA then removes the EMG.
- Poor signal Fix the signal before interpreting.
- Cautery Press and hold; the transient artefact shows only while held, then reverts.
In linked mode, the Depth slider moves the picture from awake to deep. A Transition control ramps depth changes over Instant, 30 s, 1 min, 2 min or 5 min — an educational state movement, not a drug model. It applies to the slider and to the depth presets (Adequate GA, Too light, Profound depth, Burst suppression, Isoelectric), whose depth then moves over 30 s out of the box rather than snapping; pick Instant when you want it set at once. Only depth ramps — a preset's signal conditions land immediately, so tapping Profound depth after EMG/NMBA clears the EMG bias at once while the depth eases across. The signal-condition presets — Poor signal, EMG/NMBA, Cautery — always apply immediately.
Signal conditions and artefacts
The Signal state control stages the device/electrode condition, each with a matching trace artefact:
- Good Clean signal.
- Poor contact Wander and electrode pops; the index freezes on the last trustworthy value.
- Movement Blink-shaped excursions; the index is invalid while it lasts, and the device re-acquires the signal afterwards.
- Mains 50Hz / Mains 60Hz Line interference; the index freezes.
- Lead off An absent, flat trace; the index is invalid.
Cautery is not on this control — it is the press-and-hold Cautery preset, because the artefact is a transient you hold rather than a state you leave set. It dominates the trace with broadband noise and invalidates the index for as long as you hold it.
Distinguish a near-isoelectric trace with good contact — a real, trustworthy low-activity signal — from an absent or invalid one (lead-off or cautery). The first is a genuine deep picture; the second is a technical fault, and must never be read as a low index.
Reading the display
- Raw trace on an explicit microvolt scale with a dotted zero baseline.
- Processed index, 0–100, shown against a common reference band — a reference, not a universal safe range.
- SQI (signal quality), EMG, suppression ratio and SEF95, each shown as
--when it is not yet valid rather than as a confident guess. - A short local trend of the displayed values — and, in the same slot, an optional density spectral array (DSA): a frequency-over-time colour strip of the spectrum SEF95 is read from. Brighter means more power at that frequency; it fills from the right and starts empty on connect (it never invents history you did not watch), and an interrupted signal is greyed rather than drawn as spectrum. It is a generic depth-and-suppression view — useful for seeing the spectrum slow with depth and burst suppression appear — and it deliberately does not reproduce anaesthetic-agent-specific patterns.
- Device-local smoothing (10 / 15 / 30 s) and an amplitude scale live on the monitor, and reset when it reloads. Smoothing changes the displayed response time, not the patient — a longer window simply reacts more slowly.
Advisories
Set the limits on the controller — Index low, Index high and Suppression high. They are room-shared and scenario-saveable. The monitor then shows two separate families of message, and the difference between them is the lesson:
When the signal is invalid the monitor announces the fault and raises no index advisory. An untrustworthy signal can never become a low-index alarm — and the advisory text never diagnoses awareness or depth, only that a limit was crossed.
Presenters and observers
- A presenter switches a monitor to the processed-EEG surface to show the class; the same tab consumes one monitor slot, so switching surface does not open a second connection.
- Read-only observers follow the presented surface — they see the trace, the index and the same displayed state, with nothing to operate.
Mode, depth, device modifiers, manual outputs, advisory limits. Saved into a scenario step.
Smoothing, amplitude scale, the Trend/DSA choice, alarm silence. Deliberately not saved into a scenario — and not kept across a reload either: each monitor returns to its defaults.
Scenarios and restoration
- A scenario step can capture the room's processed-EEG state — mode, the depth source, device modifiers, manual outputs and the advisory limits — and restore it as the step runs.
- Device-local display preferences are deliberately not saved into the step, and they do not survive a reload — a reloaded monitor comes back at 50 µV, 15 s smoothing, Trend, unsilenced (see the split above).
- After a reconnect or reload, re-check the presented surface and the displayed values before you rely on them for teaching — resume the picture rather than assuming it carried over.
Teaching examples
- Stable anaesthetic — the trace, index and reference band agree.
- EMG confound — muscle activity lifts the index while depth is unchanged.
- Burst suppression — activity alternates with suppression and the ratio rises.
- Poor signal — a degraded signal holds the number rather than trusting it.
Each is selectable in the interactive examples on the module overview.
Troubleshooting
- The module seems inactive
- Open the controller's EEG tab and tap Enable processed EEG, or apply a preset. Opening the tab alone does not start it.
- No valid index (the number reads
--) - Check the signal state — Lead off, Movement or Cautery invalidate the index; Poor contact and mains interference freeze it. Restore a good signal to present a live value.
- The index moves unexpectedly
- Check for high EMG (it falsely elevates the index) and whether the picture is in Manual mode, where outputs are set independently. Device-local smoothing also changes how quickly the number responds.
- The presenter and an observer show different surfaces
- The presenter chooses the surface; make sure the monitor you intend to present from has switched to the processed-EEG surface, and give observers a moment to follow.
- Remember that view preferences are device-local, so smoothing or scale can differ between screens without being a fault.
- Performance on an older device
- Close other tabs, and present from the most capable screen available. The surface pauses its own drawing when it is off-screen or the tab is hidden.
- After a reconnect or reload
- Re-open the presented surface and confirm the displayed values before relying on them.
Limits
- The module is not a reproduction of any proprietary clinical algorithm.
- It is not a diagnostic or treatment device.
- A number here is not proof of awareness, unconsciousness or anaesthetic adequacy.
- It is not yet a multi-vendor device-familiarisation suite; it is one vendor-neutral, single-index teaching model.
- Behaviour is adult-modelled. In a paediatric room the module still runs the adult model (at paediatric rates and context) — offered as an informed instructor choice, not validated for paediatric patients. The caveat is instructor-held and debriefed; it is not shown on the learner monitor.
If the patient arrests, the depth surface keeps showing whatever you set — it does not fall to isoelectric by itself, and it does not come back by itself after ROSC. That is deliberate rather than an oversight: the fall is straightforward, but how a brain's electrical activity returns after an arrest is genuinely contested, and NAP will not invent that curve on your behalf. You drive it. Isoelectric stages the arrest picture in one tap, and what returns after ROSC — and how fast — is your teaching choice.