5.4 Quantitative EEG (qEEG) Trends & ICU Trending Tools
Key Takeaways
- Quantitative EEG (qEEG) applies digital signal processing and Fast Fourier Transform (FFT) to convert multi-channel raw voltage data into time-compressed graphical trends (DSA, CSA, Rhythmicity Spectrogram, aEEG, ADR, Asymmetry Index, BSR) over 1 to 24-hour display epochs.
- Electrographic seizures manifest on Density Spectral Array (DSA) and spectrogram trends as distinctive 'flame', 'candlestick', or 'wedge' signatures reflecting rapid surges in total spectral power accompanied by dynamic upward and downward frequency evolution.
- In aneurysmal subarachnoid hemorrhage (aSAH), continuous monitoring of the Alpha-Delta Ratio (ADR) and Relative Alpha Variability (RAV) detects Delayed Cerebral Ischemia (DCI) secondary to vasospasm, with sustained >10–20% declines preceding angiographic vasospasm or clinical stroke by 24–48 hours.
- Amplitude-Integrated EEG (aEEG) applies filtering, rectification, and asymmetric logarithmic compression to display upper and lower voltage envelopes, facilitating rapid assessment of background continuity, sleep-wake cycling, and burst suppression depth.
- The Golden Rule of qEEG mandates that quantitative trends are screening and detection tools; every trend abnormality (suspected seizure, sudden asymmetry shift, burst suppression alteration) MUST be verified by opening and inspecting the corresponding raw, uncompressed EEG page and synchronized video to rule out physiological and mechanical artifacts.
5.4 Quantitative EEG (qEEG) Trends & ICU Trending Tools
Continuous EEG (cEEG) in the Intensive Care Unit generates vast amounts of data—typically 24 to 48 hours of continuous 21-channel digital recordings per patient per day. Reviewing thousands of pages of raw waveforms in real time is clinically impractical for continuous bedside surveillance. Quantitative EEG (qEEG) solves this challenge by applying digital signal processing algorithms to transform multi-channel raw EEG signals into intuitive, time-compressed graphical trends.
By compressing hours of EEG data into a single display window, qEEG enables clinicians and technologists to instantly identify paroxysmal seizures, track the depth of sedative-induced burst suppression, detect evolving hemispheric asymmetries, and identify early cortical ischemia in aneurysmal subarachnoid hemorrhage (aSAH).
1. Principles of Digital Signal Processing & FFT
qEEG trends rely on mathematical transformations that convert analog-derived time-series voltage data into frequency and power domain representations.
+---------------------------------------------------------------------------------------------------------+
| qEEG DIGITAL SIGNAL PROCESSING PIPELINE |
| |
| [1] Raw Multi-Channel EEG [2] Epoching & Windowing [3] Fast Fourier Transform (FFT) |
| +--------------------------+ +--------------------------+ +---------------------------------+ |
| | Multi-lead time series | --> | Divided into 2–4 sec | --> | Decomposes complex wave into | |
| | voltage data (uV vs time)| | epochs; Hanning window | | discrete sine/cosine components | |
| +--------------------------+ +--------------------------+ +---------------------------------+ |
| | |
| v |
| [5] Graphical Trend Display [4] Feature Extraction [3b] Power Spectral Density (PSD) |
| +--------------------------+ +--------------------------+ +---------------------------------+ |
| | Time-compressed trends | <-- | - Total Power, ADR, RAV | <-- | Power (uV^2/Hz) calculated for: | |
| | (1 hr, 4 hr, 12 hr, 24 hr| | - aEEG, Rhythmicity, ASI | | Delta, Theta, Alpha, Beta bands | |
| +--------------------------+ +--------------------------+ +---------------------------------+ |
+---------------------------------------------------------------------------------------------------------+
The Mathematical Core: Fast Fourier Transform (FFT)
- Fast Fourier Transform (FFT): An optimized algorithm that calculates the discrete Fourier transform of a signal. It decomposes complex continuous EEG epochs into their constituent sinusoidal frequency components.
- Power Spectral Density (PSD): Quantifies the absolute or relative power (measured in µV²/Hz) contained within standard frequency bands:
- Delta (δ): 0.5 to <4.0 Hz
- Theta (θ): 4.0 to <8.0 Hz
- Alpha (α): 8.0 to <13.0 Hz
- Beta (β): 13.0 to <30.0 Hz
- Gamma (γ): ≥30.0 Hz
2. Comprehensive Catalog of ICU qEEG Trend Displays
Modern ICU monitoring software suites (e.g., Persyst, Nihon Kohden, Natus, Cadwell) incorporate specialized trend packages designed for specific clinical applications.
+---------------------------------------------------------------------------------------------------------+
| CORE ICU qEEG TREND DISPLAYS & FUNCTIONS |
| |
| Trend Display Name Display Axes & Visual Format Primary ICU Clinical Use |
| +-------------------------+ +------------------------------------------+ +--------------------------+|
| | Density Spectral Array | X-axis: Time (hrs) | Visualizing seizure onset, ||
| | (DSA / Spectrogram) | Y-axis: Frequency (0–20 Hz) | background frequency shift,||
| | | Color Gradient: Spectral Power (dB / uV^2) | sedation depth ||
| +-------------------------+ +------------------------------------------+ +--------------------------+|
| | Compressed Spectral | X-axis: Frequency (Hz) | Pseudo-3D mountain plot of ||
| | Array (CSA) | Y-axis: Successive time epochs stacked | spectral peaks over time ||
| | | Z-axis (Height): Power amplitude (uV) | ||
| +-------------------------+ +------------------------------------------+ +--------------------------+|
| | Rhythmicity | X-axis: Time (hrs) | Automated detection of ||
| | Spectrogram | Y-axis: Frequency of rhythmic peak | evolving seizures and ||
| | | Color/Height: Rhythmicity Index (sharpness) | periodic patterns (LPDs) ||
| +-------------------------+ +------------------------------------------+ +--------------------------+|
| | Amplitude-Integrated | X-axis: Time (hrs) | Neonatal encephalopathy, ||
| | EEG (aEEG) | Y-axis: Logarithmic amplitude envelope (uV) | background continuity, ||
| | | Width: Bandwidth between min and max margins | burst suppression ratio ||
| +-------------------------+ +------------------------------------------+ +--------------------------+|
| | Alpha-to-Delta Ratio | X-axis: Time (hrs) | Early detection of Delayed ||
| | (ADR) | Y-axis: Ratio of Alpha Power / Delta Power | Cerebral Ischemia (DCI) in ||
| | | Trace: Hemispheric moving average | Subarachnoid Hemorrhage ||
| +-------------------------+ +------------------------------------------+ +--------------------------+|
| | Asymmetry Relative | X-axis: Time (hrs) | Hemispheric power mismatch ||
| | Index (ARI / ASI) | Y-axis: Left-vs-Right percentage imbalance | (focal stroke, acute hemat-||
| | | Zero line: Perfect hemispheric symmetry | oma, focal ictal onset) ||
| +-------------------------+ +------------------------------------------+ +--------------------------+|
| | Burst Suppression | X-axis: Time (hrs) | Pharmacological sedation ||
| | Ratio (BSR / SR) | Y-axis: Percentage suppression (0–100%) | titration in refractory ||
| | | Trace: Suppression percentage per minute | status epilepticus ||
| +-------------------------+ +------------------------------------------+ +--------------------------+|
+---------------------------------------------------------------------------------------------------------+
Density Spectral Array (DSA / Spectrogram)
The DSA is the foundational trend display in critical care neurophysiology. It generates a 2D color-coded heat map where time runs horizontally, frequency runs vertically (0 to 20 Hz), and power is represented by a color gradient (cool blue/black = low power; warm yellow/orange/red = high power).
[ DENSITY SPECTRAL ARRAY (DSA) SEIZURE SIGNATURE ]
Frequency
(Hz)
20 +----------------------------------------------------------------+
| /\ |
15 | / \ <-- SEIZURE 'FLAME' |
| / /\ \ (High power surge & |
10 | / / \ \ frequency evolution)|
| / / /\ \ \ |
5 | --------------------------/ / / \ \ \---------------------- |
| [ Low-Power Delta Baseline ] [ Seizure ] [ Post-Ictal Slow] |
0 +----------------------------------------------------------------+
00:00 01:00 02:00 03:00 04:00 (Time)
Rhythmicity Spectrogram
Unlike the standard DSA (which displays all spectral power, including broadband background noise), the Rhythmicity Spectrogram calculates a mathematical Rhythmicity Index. It isolates and measures only those frequencies that exhibit narrow-band, sharp, periodic, or sinusoidal repetition. It is the most sensitive trend for highlighting subtle focal electrographic seizures and LPDs while filtering out diffuse, non-rhythmic polymorphic background slowing.
Amplitude-Integrated EEG (aEEG)
Originally developed by Maynard in the 1960s (Cerebral Function Monitor) and ubiquitous in neonatal and adult neuro-ICUs, aEEG utilizes a specialized processing chain:
- Asymmetric Bandpass Filtering: Passes frequencies between 2 and 15 Hz, attenuating <2 Hz (movement/breathing artifacts) and >15 Hz (EMG artifact).
- Rectification & Peak Detection: Extracts the continuous upper envelope (maximum peak-to-peak voltage) and lower envelope (minimum voltage).
- Logarithmic Compression: Compresses voltages non-linearly (linear from 0 to 10 µV, logarithmic from 10 to 100 µV).
+---------------------------------------------------------------------------------------------------------+
| aEEG BACKGROUND CLASSIFICATION |
| |
| Pattern Type Lower Margin (uV) Upper Margin (uV) Clinical Interpretation |
| +-----------------------+ +---------------------+ +---------------------+ +-------------------------+ |
| | Continuous (CNV) | > 5 µV | > 10 µV | Normal / Mild slowing | |
| +-----------------------+ +---------------------+ +---------------------+ +-------------------------+ |
| | Discontinuous (DNV) | < 5 µV | > 10 µV | Moderate encephalopathy | |
| +-----------------------+ +---------------------+ +---------------------+ +-------------------------+ |
| | Burst Suppression (BS)| < 5 µV (often 0–2 µV) | > 25 µV (broad band) | Severe injury / Sedation| |
| +-----------------------+ +---------------------+ +---------------------+ +-------------------------+ |
| | Continuous Low Voltage| < 5 µV | < 10 µV (narrow band) | Severe diffuse injury | |
| +-----------------------+ +---------------------+ +---------------------+ +-------------------------+ |
| | Flat / Isoelectric | < 5 µV (0–2 µV) | < 5 µV (0–3 µV) | Electrocerebral silence | |
| +-----------------------+ +---------------------+ +---------------------+ +-------------------------+ |
+---------------------------------------------------------------------------------------------------------+
Ischemia Monitoring Trends in aSAH: ADR & RAV
In patients with high-grade aneurysmal subarachnoid hemorrhage (aSAH), Delayed Cerebral Ischemia (DCI) secondary to microvascular spasm and cortical spreading depolarizations represents the primary cause of preventable secondary infarction between days 3 and 14.
- Pathophysiology: Regional cerebral hypoperfusion reduces neuronal ATP synthesis, causing selective failure of synaptic fast frequencies (alpha/beta decay) and emergence of regional polymorphic delta slowing.
- Alpha-to-Delta Ratio (ADR): Calculated as
(Absolute Alpha Power [8–13 Hz]) / (Absolute Delta Power [1–4 Hz]). - Relative Alpha Variability (RAV): Measures the standard deviation of relative alpha power over rolling 1- to 2-hour windows (normal awake cortex exhibits high variability that fluctuates with vigilance; ischemic cortex loses variability).
- Diagnostic Alarm Threshold: A sustained relative drop >10% to 20% in ADR or RAV over 4 to 6 consecutive hours over a specific hemispheric vascular territory strongly correlates with impending DCI, often preceding angiographic vasospasm or clinical neurological deficits by 24 to 48 hours.
3. Artifact Recognition on qEEG Panels
Because qEEG algorithms mathematically process all electrical potentials without innate clinical intelligence, physiological and mechanical artifacts can distort trend lines and mimic cerebral pathology.
+---------------------------------------------------------------------------------------------------------+
| qEEG ARTIFACT DIFFERENTIATION MATRIX |
| |
| Artifact Source Visual Trend Appearance on qEEG Distinguishing Raw EEG Features |
| +-----------------------+ +----------------------------------------+ +--------------------------------+ |
| | Endotracheal | High-intensity broadband vertical flash | Violent high-amplitude mixed | |
| | Suctioning | spanning all frequencies (0–20 Hz) across| muscle, movement, and electrode| |
| | | ALL channels simultaneously | pop potentials in all leads | |
| +-----------------------+ +----------------------------------------+ +--------------------------------+ |
| | Chest Physiotherapy / | Monotonous, rigid horizontal line | Rhythmic 3–5 Hz mechanical | |
| | Percussion Bed | (typically 3–5 Hz) on DSA and Rhythmicity| tapping artifacts without | |
| | | trend; mimics electrographic status | electrographic evolution | |
| +-----------------------+ +----------------------------------------+ +--------------------------------+ |
| | IV Infusion Pump / | Extremely sharp, ultra-narrow harmonic | Rhythmic brief electrostatic | |
| | Drip Artifact | spikes at exact drip rate (e.g., 1 Hz) | spikes time-locked to infusion | |
| +-----------------------+ +----------------------------------------+ +--------------------------------+ |
| | Electrode Pop / High | Isolated extreme high-amplitude surge | Classic rapid-rise, exponential| |
| | Impedance | restricted to a single channel; skews ASI| decay transient on single lead | |
| +-----------------------+ +----------------------------------------+ +--------------------------------+ |
| | Ventilator Shivering/ | Continuous low-frequency delta sway or | High-frequency 20–35 Hz tremor | |
| | Water Condensation | prominent fast-frequency envelope surges | baseline or rhythmic tubing pop| |
+---------------------------------------------------------------------------------------------------------+
4. The Golden Rule of qEEG & Technologist Workflow
[!IMPORTANT] THE GOLDEN RULE OF qEEG: qEEG trends are screening and detection tools; they are NEVER diagnostic in isolation. Every trend change, suspected seizure candle, sudden asymmetry shift, or burst suppression alteration MUST be confirmed by opening and visually reviewing the synchronized raw EEG waveforms and synchronized bedside video before notifying clinical teams or taking therapeutic action.
+---------------------------------------------------------------------------------------------------------+
| STANDARDIZED qEEG REVIEW PROTOCOL |
| |
| STEP 1: Trend Survey (Macro-Review) |
| - Set display window to 4-hour or 8-hour view |
| - Scan DSA, Rhythmicity Spectrogram, aEEG, and ADR for sharp deflections, flames, or asymmetry shifts |
| |
| STEP 2: Targeted Raw Waveform Verification (Micro-Review) |
| - Click directly on the peak of any suspicious trend deflection |
| - Open the synchronized raw EEG page (10–15 sec/page view) in standard longitudinal bipolar montage |
| - Determine if deflection represents true electrographic seizure, periodic discharge, or artifact |
| |
| STEP 3: Synchronized Video Cross-Check |
| - Check time-locked video at the event onset to inspect patient behavior, nursing care, or movement |
| |
| STEP 4: Trend Annotation & Escalation |
| - Annotate verified seizures, clinical changes, or identified artifact sources directly on the record |
+---------------------------------------------------------------------------------------------------------+
5. Clinical Traps & High-Yield Exam Pitfalls
[!CAUTION] Critical qEEG Traps for the CLTM Candidate:
- Trap 1: The 'Percussion Bed Status Epilepticus' Trap: Automated chest percussion therapy creates a continuous, high-amplitude 4 Hz rhythm that appears on the Rhythmicity Spectrogram as an alarming 'status epilepticus' block. Always cross-check the raw EEG and video before sounding an emergency alarm.
- Trap 2: Asymmetry Caused by Single-Electrode Faults: A dried electrode or high contact impedance (>10 kΩ) on a single lead (e.g., F3) will inject 60 Hz line noise or artifactual slow drift, causing the Asymmetry Index (ASI) to falsely indicate severe hemispheric pathology. Always verify electrode impedances before interpreting asymmetry trends.
- Trap 3: Missing Low-Amplitude Subclinical Seizures on aEEG: Low-voltage focal seizures (e.g., temporal lobe seizures <30 µV) may not generate sufficient broadband power to produce an obvious rise on aEEG or DSA. Technologists must combine trend screening with regular visual inspection of the raw EEG.
A technologist reviewing a 4-hour ICU qEEG trending panel notices a recurrent pattern on the Density Spectral Array (DSA) characterized by sudden, triangular, high-power surges rising from a 2 Hz delta baseline up to 14 Hz over 60 seconds, accompanied by a sharp vertical spike on the Rhythmicity Spectrogram. What does this characteristic 'flame' or 'candlestick' signature represent?
While monitoring an ICU patient on continuous qEEG, the technologist observes an alarming, high-power rhythmic band at 4.0 Hz on the Rhythmicity Spectrogram spanning 30 minutes. Upon opening the synchronized raw EEG and video, the technologist observes a continuous 4 Hz mechanical deflection in all channels caused by an active automated pneumatic chest percussion therapy vest, with underlying normal waking background. What is the correct immediate action?
A patient with high-grade Fisher Grade 3 aneurysmal subarachnoid hemorrhage (aSAH) is monitored with continuous cEEG and quantitative trending. On post-bleed day 7, the technologist observes a progressive, sustained 25% decline in the Alpha-to-Delta Ratio (ADR) and a marked reduction in Relative Alpha Variability (RAV) over the left middle cerebral territory over a 6-hour period. What clinical complication does this trend pattern signify?
On an Amplitude-Integrated EEG (aEEG) trend display in an adult ICU patient receiving therapeutic propofol infusions, the lower margin of the envelope is at 1.5 µV while the upper margin exhibits wide oscillations reaching 35 µV, creating a broad, dark vertical band on the trend. How is this aEEG background classified?