Notes

Slide Show

Outline

1	Efficient Design of Event-related fMRI Experiments with M-sequences Giedrius Buracas Geoff Boynton SNL-B, Salk Institute
2	Rapid Event-related vs. Block Design ER design is preferred for estimation of Hemodynamic Response (HDR) function, block design is preferred for detection of activation (Liu et al, 2001). Randomized ER design may resolve activation induced by individual cognitive events, that are lumped together during block design experiments. ER design might be necessary for studies that have to minimize habituation and expectation effects and for studies that are addressing these factors.
3	ER-fMRI responses
4	HDR estimation
5	Maximum likelihood estimation (e.g. Dale, 1999)
6	Estimation efficiency (expected accuracy of HDR estimate)
7	Strategies for maximizing HDR estimation efficiency
8	Conditions of Maximum Estimation Efficiency
9	Our strategy
10	What are m-sequences
11	Stochastic vs. m-sequence-governed event vectors
12	M-sequence efficiency is hard to reach by random search
13	Efficiency of random vs. m-sequences
14	m-sequences for overlapping events
15	m-sequences for non-overlapping events
16	Real data
17	Recipe
18	p.s.