First Run

This guide walks you through using EyeSort for the first time with the sample dataset.

Before You Start

1. Ensure EyeSort is installed correctly
2. Download the sample dataset
3. Extract the sample dataset to a known location
4. Launch EEGLAB in MATLAB

The EyeSort Workflow

EyeSort follows a sequential workflow:

1. Load EEG Dataset(s) - Import your synchronized EEG + eye-tracking data
2. Setup Interest Areas - Define text or pixel-based regions
3. Eye-Tracking Event Labeling - Apply labeling criteria to fixations/saccades
4. Generate BDF File - Create BINLISTER bin descriptor file
5. Save Labeled Dataset - Export for ERP analysis

Step 1: Load Dataset

In EEGLAB, go to: EyeSort → 1. Load EEG Dataset(s)

You'll see two options:

Option A: Single Dataset Mode

• Use "Browse" to select a single .set file
• The dataset loads into EEGLAB immediately
• Good for testing or single-file workflows

Option B: Batch Mode

• Select an input directory containing multiple .set files
• Select an output directory for processed files
• EyeSort validates files and loads the first dataset for preview
• All subsequent steps will process files one-by-one

For your first run, use Single Dataset Mode with the sample file.

Step 2: Setup Interest Areas

Go to: EyeSort → 2. Setup Interest Areas → Text-Based Sentence Contents and Interest Areas

This opens a configuration GUI where you'll specify:

Key Settings for Sample Dataset:

1. Browse for IA file - Select the tab-delimited interest area file
2. Sentence offset - Horizontal pixel offset (e.g., 281)
3. Pixels per character - Character width in pixels (e.g., 14)
4. Number of regions - How many interest areas per trial
5. Region names - Comma-separated list (e.g., "Region1, Region2, Region3")
6. Trigger codes:
   • Trial start/end (e.g., S254, S255)
   • Condition triggers (e.g., S211, S213, S221, S223)
   • Item triggers (supports ranges: S1:S112)

What This Step Does:

EyeSort will:

• Read your IA file and compute pixel boundaries for each word/region
• Match trials using trigger codes
• Assign region membership to each fixation
• Determine pass structure (first pass, regressions, etc.)
• Add metadata to EEG.event and EEG structure

Tip: Click "Save Configuration" to reuse these settings later!

Step 3: Label Events

Go to: EyeSort → 3. Eye-Tracking Event Labeling

Now you can apply labeling rules. For example:

Example Label: "First Fixation on Region 2"

1. Time-locked region: Select "Region2"
2. Pass: Select "First"
3. Fixation type: Select "First of Multiple"
4. Label description: Enter "First fixation on Region 2 during first pass"
5. Click "Add Label"

Each label increments a counter and assigns a 6-digit code: CCRRLL

• CC = Condition code (2 digits)
• RR = Region code (2 digits)
• LL = Label number (01-99)

Try Multiple Labels:

Create labels for different patterns:

• First fixations vs. subsequent fixations
• Forward saccades vs. regressions
• First pass vs. second pass
• Different regions

Tip: Save your label configuration for reuse across datasets!

Step 4: Generate BDF File

Go to: EyeSort → Generate BINLISTER BDF File

EyeSort will:

• Scan all labeled events
• Group them by condition and region
• Generate a BINLISTER-compatible BDF file (default: eyesort_bins.txt)
• Use your label descriptions for bin text

This BDF file can be used with ERPLAB's pop_binlister() function.

Step 5: Save Dataset

Go to: EyeSort → Save Labeled Dataset

Choose a filename (e.g., sample_eyesort_labeled.set) and save.

Your dataset now contains:

• Original data + events
• Labeled event codes
• Region/pass/fixation metadata
• Configuration information

Verify Your Work

Open the saved dataset in EEGLAB and inspect the events:

EEG = pop_loadset('sample_eyesort_labeled.set');
% View first few events
EEG.event(1:10)

You should see:

• Event types with 6-digit codes (e.g., 011101)
• original_type field preserving original event names
• eyesort_* fields with metadata

Next Steps

Now that you've completed your first run:

1. Try the full Quickstart Tutorial with your own data
2. Learn about Interest Areas
3. Understand Pass Structure
4. Explore Batch Processing for multiple files

Common First-Time Issues

"EEG data is not properly processed with region information"

• You skipped Step 2 (Setup Interest Areas)
• Run the Text IA setup first

"No labeled events found"

• Your labeling criteria might be too strict
• Try relaxing constraints (e.g., use "Any" for pass or fixation type)

Field name mismatches

• Enable "Change default event field names" in Text IA GUI
• Specify your dataset's exact field names