Requirements

Download & Install

Go to bcinexus.io/download, select your OS, and download the installer (~800 MB).

Windows

1. Run BCILattice-Setup.exe
2. Accept the license agreement and choose an install location (default: C:\Program Files\BCILattice\)
3. Allow PostgreSQL installation when prompted by Windows Firewall
4. Launch BCILattice, check the bottom status bar shows Server: Running

macOS

1. Open BCILattice.dmg and drag BCILattice.app to Applications
2. First launch: right-click → Open (macOS Gatekeeper bypass, only needed once)
3. Allow BCILattice through the Firewall when prompted (for the local backend)
4. Apple Silicon (M1/M2/M3) is supported natively with Metal acceleration

Linux

chmod +x BCILattice-installer.AppImage
./BCILattice-installer.AppImage

Works on any modern Linux distro (glibc 2.31+). No root access required.

Create an Account

BCILattice works fully offline. A free BCINexus account is optional but unlocks cloud backup, 340+ community pipelines, and team collaboration.

1. Click Sign In / Create Account in BCILattice top-right
2. Or register at bcinexus.io/register
3. Verify your email, then sign in inside BCILattice

You can skip this and start immediately, add an account later in Settings.

Import Your First Dataset

Don't have data yet? Download a free 10-subject EEG motor imagery sample at bcinexus.io/sample-data.

Single file:

1. Click Data Manager in the left sidebar
2. Click Import Files → select your EEG/fNIRS/EMG file
3. BCILattice previews signal + metadata, confirm label column
4. Click Add to Session

Multi-subject folder:

1. Click Import Folder → select root directory
2. Set folder depth (1–5 levels) → BCILattice scans recursively
3. Set session name and label config → click Import All

Configure Preprocessing

Click Preprocessing in the left sidebar. Quick-start settings for EEG:

Click Preview to see before/after, then Apply to Session.

For fNIRS: enable the one-click TDDR → OD → Beer-Lambert pipeline toggle.
For EMG: enable full-wave rectification + envelope extraction.

Design a Paradigm with NeuralFlow

Click Neural Flow. Drag blocks from the left palette onto the timeline canvas.

5-minute motor imagery paradigm:

1. Drag Start Trial block (code -1)
2. Drag Visual Cue (2 s), prompt the subject
3. Drag MI Left Hand (4 s) · MI Right Hand (4 s)
4. Drag Rest (2 s), inter-trial rest
5. Drag End Trial block (code -2)
6. Wrap steps 2–4 in a Loop Block, set Total Repeat to 20
7. Click Compile → Export → .nflow

Build & Train a Machine Learning Pipeline

Click ML Suite. Drag blocks from the catalog panel, connect ports, then compile and train.

Simple EEG classification pipeline:

1. Drag Channel Selection block, connects to your imported session
2. Drag StandardScaler → connect to Channel Selection output
3. Drag EEGNet (or LDA Classifier for a fast baseline) → connect to scaler
4. Drag accuracy_score block → connect to classifier output
5. Click Workflow Designer tab → assign subjects to the pipeline
6. Click Compile (Ctrl+Shift+B) then Train

The Batch Training Dashboard opens with live per-subject progress. Training runs on your local CPU or NVIDIA GPU.

Review Results in Experiment Hub

1. Click Experiment Hub, completed run appears as v1
2. Click any row to see per-subject accuracy, F1, AUC
3. Change a parameter in ML Suite → compile → train → results appear as v2
4. Select v1 and v2 → click Compare for side-by-side
5. Right-click → Restore to roll back any previous configuration

Generate a Report

1. Click Reports → Generate Report
2. Choose HTML or PDF
3. Report includes: session metadata, preprocessing config, NeuralFlow paradigm, per-subject results
4. HTML reports are standalone, email them directly

Keyboard shortcut: Ctrl+R / Cmd+R

Explore Community Pipelines

1. Go to bcinexus.io/pipelines
2. Browse 340+ pipelines by modality or task (motor imagery, P300, SSVEP…)
3. Click Import into BCILattice, BCILattice opens and imports the pipeline automatically

Keyboard Shortcuts

Troubleshooting