A Fast Algorithm for Constructing Efficient Event-Related

7 We propose a novel, efficient approach for obtaining high-quality experimental designs for 8 event-related functional magnetic resonance imaging (ER-fMRI). Our approach combines a 9 greedy hillclimbing algorithm and a cyclic permutation method. When searching for optimal 10 ER-fMRI designs, the proposed approach focuses only on a promising restricted class of 11 designs with equal frequency of occurrence across stimulus types. The computational time 12 is significantly reduced. We also demonstrate that our proposed approach is very efficient in 13 achieving designs that are robust against mis-specification of the error autocorrelation. 14 keywords: A-optimality; cyclic permutation; D-optimality; genetic algorithm; hillclimb15 ing; maximin design 16