Hi,

We are using the SMI 50 Hz eye tracker in our MEG and
MRI labs at the moment. We are planning to upgrade to
the new 250 Hz system as soon as it becomes available.

The following only refers to the MEG system.

The IR light source can indeed just be seen by the subject,
but this was never a problem, as it is very faint. The camera
sits under the screen, at about 130 cm away from the
helmet. We never had problems with interference.

Setup never takes more than about 5-10 minutes, but we
do have some problems with droopy eyelids with 10-20%
of our subjects. This might be because a relatively high
percentage of subjects have been older lately, because
of a few dementia studies. But I do think that MEG in
general causes people to be a bit sleepy, as it is very
quiet in the MSR, and most tasks are quite boring.

We did visit the MEG lab in London a while ago to have
a look at the SR system. It is a lot faster than the SMI
system, but SMI's software seems to be more user-
friendly and easier to integrate with the acquisition
software. If 250 Hz is fast enough you might want to
go for the SMI eye tracker, otherwise the SR one is the
only option.

Both the SMI and the SR trackers were not able to
cope with the droopy eyelids. The SMI software outputs
the wrong gaze location: lower than it actually is, while
the SR system simply gives no location at all when the
pupil is partly covered. I guess no output is better than
the wrong output. SMI have told me that the latest
version of their software is better able to cope with
this, but I haven't been able to verify this yet.

We do not have any experience with other systems.

Maarten