Deconvolution of impulse response in event-related BOLD fMRI.

TLDR: Deconvolution may be effective in diminishing the hemodynamically imposed temporal blurring and may have potential applications in quantitating responses in eventrelated fMRI.

About: This article is published in NeuroImage.The article was published on 1999-04-01 and is currently open access. It has received 1343 citations till now. The article focuses on the topics: Finger tapping & Wiener deconvolution.

Figures

FIG. 1. Stimulus for experiment I. A visual cue (a blue box show nger-tapping task. The tones are square-wave modulated at 3 Hz for here are 10 trials in each scan. Images are acquired at 1-s intervals,

FIG. 2. Stimuli for experiment II, showing 1 trial of 10 total. The ones are square-wave modulated at rates between 1 and 4 Hz (D). (A) or experiment IIA, a visual cue begins 1 s before the tones. eparation between adjacent stimulus episodes Ts varies between 1 nd 5 s, with 3-Hz tone/finger-tapping rate. Each trial is 30 s long. (B nd C) For experiment IIB, a nonmodulated 1-s-long auditory lerting cue begins 1 s before the finger-tapping tones. Two or four 2-s timuli are delivered with 4 s spacing between episodes. In some xperiments the pacing rates Fi were identical (see Table 1). The IDI pitch of both the alerting tones and the pacing tones is linearly elated to Fi: for F 5 1 Hz, P 5 66, A 5 58; for Fi 5 2 Hz, P 5 70, A 5 2; for Fi 5 3 Hz, P 5 72, A 5 64; for Fi 5 4 Hz, P 5 74, A 5 66. Images re acquired at 1-s intervals, with t 5 0 at the beginning of the

FIG. 10. Comparison of calculated response using balloon mode alculations use modified Buxton’s model.

FIG. 9. Effect of noise parameter on deconvolution, using m(t) data from Fig. 7G. (A) h(t) from Fig. 7F, 3 Hz; (B) m(t); (C) spectrum of h(t); D) deconvolution results for N0 5 10 (x), 20 (o), 40 (*). (E) h(t) from averaged response, Fig. 3A; (F–H) as in (B–D).