Installation & Setup

The Coherence Workstation runs as a desktop application on your computer. Your EEG data stays local—nothing is uploaded to a cloud service, nothing leaves your machine unless you explicitly choose to use the AI Research Assistant, which sends anonymized analysis context (never raw recordings) to generate its responses. Setting up means installing the application and creating a workspace where your subjects and sessions will live.

What You Need

You’ll need a Mac or Windows computer with enough horsepower to handle the signal processing. The Coherence Workstation runs ICA decomposition, source localization, connectivity estimation, and time-frequency analysis on every recording—these are computationally intensive operations, the same ones that make EEGLAB sessions crawl on underpowered machines.

macOS—Apple Silicon (M1 or later) or Intel. macOS 12 Monterey or later.

Windows—Windows 10 or later, 64-bit.

RAM—16 GB recommended. ICA decomposition and source estimation are memory-hungry. 8 GB will work for shorter recordings, but you’ll notice the difference.

Storage—Your recordings, processed stage files, and reports all live on disk. A typical fully-processed session produces 50–100 MB of analysis output. Plan accordingly if you’re building a library of sessions over time.

Internet—Only required for the AI Research Assistant. All signal processing, analysis, and report generation happen locally, offline. If you never use the AI features, you never need a network connection.

Getting the Application

Coherence Workstation is currently in early access. We set up each new clinician directly—you’ll receive the application and a walkthrough from Peak Mind. This isn’t a limitation we’re hiding behind. It’s intentional.

Early access means we can make sure your first experience actually works. We catch issues specific to your recording setup, your amplifier’s file format, your channel configuration—before you hit them alone. We incorporate your feedback in real time. And when something doesn’t behave the way you expect, you have a direct line to the people who built it.

When the application is ready for general release, this page will have a download link and a standard installation walkthrough. For now, the guided setup is the product.

Creating Your Workspace

A workspace is a folder on your computer that holds everything: your subjects, their sessions, the raw recordings you import, and all the processed analysis output the pipeline generates. Think of it as your practice’s clinical EEG library.

When you first open the Coherence Workstation, you’ll create a new workspace or open an existing one. Choose a location with enough storage for the sessions you plan to accumulate—an external drive works fine if your internal storage is tight.

Inside your workspace, the application organizes data automatically. Each subject gets a folder. Each session within a subject gets its own subfolder, dated. Raw recordings, processed stage files, and generated reports all live in their expected places. You don’t need to manage this structure manually—the application handles it.

Workspace Settings

Your workspace carries default settings that apply to every new session you create. These include pipeline parameters—filter bands, notch frequency, ICA thresholds, which analysis modules to run—and preferences like your AI provider configuration and display theme.

The defaults are sensible starting points. You can adjust them at any time through Settings → Pipeline, and you can override them per-session when you need to. But for your first recording, the defaults will serve you well. The Your First Analysis walkthrough covers the settings you’ll actually interact with.

Organizing Your Recordings

The Coherence Workstation expects a folder containing your recording files for each session. When you create a new session and point it to a folder, the application scans for recognized file types—NFX files from a Neurofield Q20/Q21, EEGLAB .set files, EDF recordings—and automatically identifies what it finds: resting eyes-open, resting eyes-closed, ERP task data.

You don’t need to rename files, reorganize directories, or follow a specific naming convention. The discovery process handles classification. If it gets something wrong, you can correct it before processing begins.

For details on what file formats the workstation supports and what to expect from different recording systems, see File Formats.

You’re Ready

Your workspace is set up, your recordings are accessible, and the application knows where to find them. The next step is running your first analysis—loading a recording, watching the pipeline process it, reviewing the results, and opening your first structured clinical reading.

Head to Your First Analysis to walk through it.