1. Eye Tracking and its Application in MRI and EEG Settings
Marcus Johnson, Ph.D
SR Research Ltd.
Toronto - Ottawa, Canada

2. Agenda Discussion of eye tracking hardware components
System cabling and integration for noise/artifact minimization
Synchronization options
Discussion of Recording/Messaging Options

3. System Hardware

4. Desktop Mount Useful in EEG Environment
Remote Mode with Sticker on EEG Cap
Chinrest Mode for More Precise Recording – One concern is Chinrest

5. Long Range Mount Camera Head Infrared Illuminator with Focusing Lens
Camera Lens
Different Lenses (75mm,50mm,35mm) for
Different Distances (overall range: cm)
Required for MRI, Potentially Useful with EEG

6. Typical MRI Configuration

7. MRI Mounting Options - Tripod
(Picture from MEG setup)
Difficult to find tripods with no ferromagnetic metals in them

8. MRI Mounting Options – Screen Mount
No ferromagnetic
metals
Flexible positioning

9. MRI Mounting Options – Screen Mount
No ferromagnetic
metals
Flexible positioning
Works with most MRIs

10. MRI Mounting Options – Trio Tray Mount
No ferromagnetic
Metals
Rests on Trio’s
bed rails
Optional built-in screen
Works with Siemens
3T Trio

11. MRI Head Coil Mirrors Allows subject to see screen and for eye
tracker to see subject
Best to use front-surfaced
mirrors to prevent ghosting
and secondary reflections

12. Long Range System Cabling

13. Long Range System Cabling
Camera Base Box on Side of Host PC
Fiber Optic Data Cable
Power Options
Lemo Connectors
(to Camera/Illum.)
Patch Panel Connection Options (DB9 or BNC)
Battery Option

14. Long Range System Cabling

15. Long Range System Cabling
No detectable interference with proper cabling
(Graph from system use in MEG)

16. System Synchronization

17. System Synchronization
Pre-Experiment activities
Check Tracker Settings (via Set Options Screen and sending commands from Display PC)
Participant setup
Calibration
Validation

18. MRI/Eye Tracker Synchronization
Running the experiment
Blocks
Trials
MRI sends synchronization pulse (TTL) to Display PC – Display PC in turn sends Message to Host PC
Pulse triggers Display screen onset
– on screen onset Display PC sends Message to Host PC
Pulse typically sent from MRI to Parallel Port or
USB Device
Can optionally be split and sent to Host PC Parallel Port
Parallel Port status on Host recorded on every sample
Optional drift correction/drift checking
Usually between Blocks
Can also enable online drift correction with mouse click

19. MRI/Eye Tracker Synchronization
TTL
Pulse
Display PC
Updates Screen
Message
When
Pulse
Received
Message
When
Screen
Updated
Time

20. EEG/Eye Tracker Synchronization
Running the experiment
Blocks
Trials
Display PC draws to screen – at same time sends
Message to Host PC and sync pulse (TTL) to EEG
Message/Pulse occur at screen retrace event
Pulse typically sent from Parallel Port or USB Device of
Display PC to EEG – can also be sent from Parallel Port
of Host PC
Event types can be coded (32 different pulse values)
Optional drift correction/drift checking
Usually between Trials
Can also enable online drift correction with mouse click

21. Analog Card Option for Host PC – usually used for devices like EEG
Converts Digital Data to Analog Voltages
Output to BNC connectors
Voltage range configurable
Three Channels per eye being tracked
Horizontal Position
Vertical Position
Pupil Size
Also allows for extra Digital In/Out (in addition to Parallel Port)
Quick –
Data can be inserted into empty EEG channels
Allows for easy alignment with EEG data
Dirty –
Noise is added by D/A and A/D conversion
Saccade/Fixation information is lost

22. EEG/Eye Tracker Synchronization
Display PC
Updates Screen
Message
When
Screen
Updated
TTL
When
Screen
Updated
Within
1 msec
Optional
Constant
Analog
Out to EEG
Time

23. EEG/Eye Tracker Synchronization
Display PC
Updates Screen
Message
When
Screen
Updated
About
1 msec OR
Optional
Constant
Analog
Out to EEG
TTL
from
Host PC to
EEG
Time

24. Recording and Messaging

25. Recording and Trials Messages are Not Just For Marking Stimulus Events
Data Viewer Format:
Trial Onset Messages
TRIALID 1 (TRIALID2, etc.)
Trial Event Messages
DISPLAY_1_ON
BUTTON_RECEIVED
DISPLAY_2_ON
Trial Variable Messages
!V TRIAL_VAR trial_condition condition_1
!V TRIAL_VAR reaction_time 2364
Trial Interest Area Messages
!V IAREA RECTANGLE square
Target Position Messages (for moving targets)
!V TARGET_POS Targ1 (512, 384) 1 0
!V TARGET_POS Targ1 (512, 364) 1 0
Trial Offset Messages
TRIAL_RESULT 0

26. Recording and Trials Messages are Not Just For Marking Stimulus Events
Data Viewer Format:
For full description of Data Viewer format see:
Data Viewer User Manual (Help -> Contents)
“Protocol for EyeLink Data to Viewer Integration”

27. Recording and Trials
EyeLink Recording and Analysis Messages are Independent
Scenario 1:
Start Recording
Start Trial Loop
For Each Trial:
Trial Onset Message
Stimulus Event Messages
Mark Every Event of Trials
Trial Condition Messages
Independent Variables
Behavioral (non eye-based) Dependent Measures
Trial Interest Area Messages
Trial Offset Message
End Trial Loop
Stop Recording

28. Recording and Trials
EyeLink Recording and Analysis Messages are Independent
Scenario 2 (Like Experiment Builder):
Start Trial Loop
For Each Trial:
Start Recording
Trial Onset Message
Stimulus Event Messages
Mark Every Event of Trials
Stop Recording
Trial Condition Messages
Independent Variables
Behavioral (non eye-based) Dependent Measures
Trial Interest Area Messages
Trial Offset Message
End Trial Loop

29. Recording and Trials
EyeLink Recording and Analysis Messages are Independent
Can Use Scenario 1 or 2 for Most Programming Environments
e.g., Psychtoolbox (Matlab), E-Prime, Presentation, C,
Python
If using Experiment Builder, Must Stick to Scenario 2
Many Messages Sent Automatically with Experiment Builder

30. Thank you!

31. EyeLink Support Documents
EyeLink 1000 User Manual
EyeLink 1000 Installation Guide
SR Research Experiment Builder
Windows Programmers Guide
EyeLink Data Viewer
Contact Information
Phone: /
Web:

32. EyeLink Support