NeuroDesign - Methods

About the algorithm

This toolbox is an easy-to-use GUI implementation of the genetic algorithm for design optimization. A good overview about the genetic algorithm can be found in this paper:

Wager and Nichols (2002). Optimization of experimental design in fMRI: a general framework using a genetic algorithm. NeuroImage 18, 293-309.

More recently, Ming-Hung Kao has further developed the genetic algorithm. The details can be found in the following papers. This application uses their algorithm.

Kao, Mandal, Lazar and Stufken (2009). Multi-objective optimal experimental designs for event-related fMRI studies. NeuroImage 44, 849-856.

Kao, Temkit and Wong (2014). Recent developments in optimal experimental designs for functional magnetic resonance imaging. World Journal of Radiology, 6(7), 437-445.

If you're not really interested in optimising the power or efficiency of your design, but only in making sure there are no accidental contigencies in your (random) design, take a look at this paper about pseudo-random designs using m-sequences.

Buracas and Boynton (2002). Efficient design of event-related fMRI experiments using M-sequences. NeuroImage 16, 801 – 813.

About design optimisation

Jeanette Mumford has made some really cool video's that give intuition and insight about why you should care about design optimisation:

• Estimation vs. detection
• The math behind it
• Here's a demo in matlab about what variable ITI's can do for your experiment