| |
Abstract:
Current fMRI techniques often require the timing of trial
responses to be specified in advance. A number of cognitive
paradigms exist that require participants to make responses in a
self-paced manner or lock to events that are unpredictable in time.
Here we report on the application of a self-paced event-related
fMRI design that also yields considerable response sampling for
precise temporal estimates of the hemodynamic response. 17
volunteers were imaged (whole-brain fMRI, 1.5 Tesla scanner,
asymmetric spin-echo, TR = 2.36 sec) while performing a mental
rotation task. Each trial began 750 msec after subject's response
in the previous trial (index finger key-press). Trial onsets were
thus paced by the subject, independent of scanner acquisition
timing. A direct comparison of right- and left-hand responses
showed strong activation in motor cortex, SMA, and cerebellum, both
at the group and at the single-subject level. The increased
temporal resolution obtained with the self-paced design allowed
accurate quantification of amplitude and timing estimates. Strong
correlation was observed between reaction times and both
time-to-peak (r2 = .61) and time-to-onset (r2 = .42) response
estimates. Mean temporal offsets of < 600 msec could be reliably
detected. These methods have a direct application to a variety of
cognitive tasks, and can advance the investigation of timing of
cognitive processes. Supported by McDonnell Center for Higher Brain
Function.
|
