Power and sample size calculation for neuroimaging studies by non-central random field theory

Abstract

Determining power and sample size in neuroimaging studies is a challenging task because of the massive multiple comparisons among tens of thousands of correlated voxels. To facilitate this task, we propose a power analysis method based on random field theory (RFT) by modeling signal areas within images as non-central random field. With this framework, power can be calculated for specific areas of anticipated signals within the brain while accounting for the 3D nature of signals. This framework can also be extended to visualize local variability in sensitivity as a power map and a sample size map. We validated our non-central RFT framework based on Monte-Carlo simulations. Moreover, we applied our method to a blood oxygenation level dependent (BOLD) functional magnetic resonance imaging (fMRI) data set with a small sample size in order to demonstrate its use in study planning. From the simulations, we found that our method was able to estimate power quite accurately. In the fMRI data analysis, despite the small sample size, we were able to determine power and the number of subjects required to detect signals.

Age-related changes in the functional neuroanatomy of overt speech production

“…In the study of Desmond and Glover (2002), about 12 subjects were required to achieve 80% power at the single voxel level, based on parameter estimates from a verbal working memory experiment. Similarly, the study of Hayasaka et al (2007) found that approximately 13 subjects would be required to achieve an 80% power in an auditory stimulation experiment. These papers, however, estimated sample size and statistical power for within-group analyses.…”

“…To estimate the sample size necessary to achieve an 80% power in our study, we used the results of two papers that performed simulation experiments and analyzed data of real fMRI experiments (Desmond and Glover, 2002;Hayasaka et al, 2007). In the study of Desmond and Glover (2002), about 12 subjects were required to achieve 80% power at the single voxel level, based on parameter estimates from a verbal working memory experiment.…”

A bi-ventricular cardiac atlas built from 1000+ high resolution MR images of healthy subjects and an analysis of shape and motion

“…In our case, statistical parametric mapping involves multiple hypothesis tests, which makes the population size question even more complex. (Friston et al, 1999) and (Hayasaka et al, 2007) explored this question for brain image studies and suggested that random models and statistical power analysis can be used to determine the population size.…”

Statistical Analysis for Recovery of Structure and Function from Brain Images

“…RFT correction is less conservative than the Bonferroni correction. In this work, we use the non-central F RFT to correct the multiple comparison problems (Wang & Xia, 2009;Hayasaka et al, 2007). ,,, , , (Anderson, 2008), its variance-covariance matrix can be calculated as: ,, ,…”