5 min read
Updated September 9, 2025
A thin-film EEG grid records through the skull, so there is no craniotomy and no cortical exposure. That removes most of the surgical risk. It concentrates the difficulty in two places: aligning the array to the anatomy, and keeping it flat against bone for the life of the study.
Research use only. NeuroNexus products and the procedures described here are for non-clinical research use, and must not be used in human or veterinary medical procedures.
This guide describes an approach, not a protocol. Adapt it to your species, your preparation and your institutional requirements. All procedures run under IACUC approval.
EEG or ECoG#
The choice is made before surgery and it is not reversible.
| EEG grid | ECoG grid | |
|---|---|---|
| Placement | Skull surface | Cortical or dural surface |
| Craniotomy | None | Required |
| Amplitude | Microvolts | Higher; better spike-band content |
| Spatial resolution | Broader | Finer |
| Chronic risk | Low | Higher |
EEG is the right choice for sleep and wake staging, seizure detection, cortical rhythms and long-duration monitoring. Coverage and implant survival matter more there than resolution.
Before surgery#
Anesthesia and analgesia per institutional guidelines, with multi-day post-operative analgesia planned.
Aseptic technique. This is survival surgery.
Instruments. Fine forceps, scissors, hemostats, scalpels, needle holders; a high-speed drill with burrs for the anchor screws; a micropipette or blunt needle for saline. Add a tool suited to handling a flexible array — a polyimide grid creases if it is picked up by an edge.
Implant materials. Dental cement or a biocompatible adhesive, bone screws, and a protective housing for the connector.
Rig. Stereotaxic frame and a dissecting microscope with a cold light source.
Surgery#
Site preparation#
Secure the subject in the frame. Incise the scalp and expose the skull over the target area. Remove the periosteum and rinse with saline.
Dry the bone before placement. The array couples through the skull, and a film of fluid under the grid becomes a shunt between adjacent sites.
Smooth any ridges. A grid tented over an irregularity loses contact at the sites either side of it.
Placement#
Align to bregma. Bregma is the only landmark you have, and every claim you make about which cortical area a channel covers rests on it. Mark the intended grid outline before the array touches the skull.
Lay the array down from one edge, rolling it into contact rather than dropping it flat. This is what keeps air out from under the sites.
Do not reposition by sliding. Lift, re-wet and re-place. Sliding a polyimide array across bone abrades the sites.
Fixation#
Screws first. Bone screws carry the mechanical load and provide the ground and reference anchors. Place them clear of the grid outline so the cement can flow around both.
Cement in thin layers. Cover the array edges and the cable exit, and build up rather than pouring. Leave the connector accessible at an angle a cable can reach without levering.
Reference and ground#
Catalog EEG grids carry 6, 16, 30 or 32 sites.
16-site grids. With a separate reference, connect the reference wire to a skull location or bone screw away from the recording area. Without one, connect the grid reference input to a bone screw.
30- and 32-site grids. With the internal reference, connect the designated wire to a bone screw. To reference externally, follow the montage your analysis expects.
Reference choice determines what your montage means. Decide it with the analysis, not at the bench. See Thin Film Grids: EEG Referencing Techniques.
Recovery#
Analgesia and monitoring per institutional guidelines. House individually with nesting material that cannot catch the connector.
Record a baseline before the experimental manipulation begins. EEG is interpreted against the animal's own baseline, and a study without one has nothing to compare against.
Recording considerations#
Amplitude. Skull-surface EEG is a microvolt signal. It needs high gain and a low-noise front end, and it is unforgiving of a poor ground.
Frequency bands. Delta 0.5–4 Hz, theta 4–8 Hz, alpha 8–13 Hz, beta 13–30 Hz, gamma 30–100 Hz. Band definitions vary between species and between literatures; state the ones you used.
Sampling. 250–1,000 Hz covers the standard bands. Anti-alias filter before decimation.
Artifacts. EMG from temporalis and neck muscles is the dominant contaminant in rodent EEG, and it sits in the same band as gamma. Place sites away from muscle where the anatomy allows, and verify any gamma result against a movement channel.
Line noise. Ground and shield properly rather than notching after the fact. A notch filter at 50 or 60 Hz removes the artifact and the real activity underneath it.
Troubleshooting#
One region of the array is quiet — usually contact, not electronics. Check for tenting or a fluid layer at that part of the skull.
All channels noisy — check ground continuity and the reference screw first.
Impedance rising over weeks — cement lifting at an edge. Inspect the perimeter.
Support#
Technical questions: support@neuronexus.com · +1.734.913.8858 Full protocol: EEG Arrays Guideline ↗