The purpose of this project is to simulate an ecologically valid fmri meta-analysis.
To be able to test meta-analytical tools and methods it is important to have a meta-analysis set that resembles a real meta-analysis in within- and between "study" variance but in the meantime does not suffer from the same lack of data as fMRI meta-analyses. In this project we aim to develop such a meta-analysis set and test it's within- and between "study" variance and the influence of family ties on the variability.
We use a meta-analysis on pain which can be downloaded from neurovault (http://neurovault.org/collections/1425/) (https://www.nature.com/articles/sdata2016102). If the folder with the data is put into the working directory with the code, everything should work fine.
- Read in t-maps
- Compute effect sizes
- Calculate within- and between study variance
- Pool HCP studies
- Compute within- and between study variance of HCP meta-analysis
- Determine other parameters from both meta-analyses (number of peaks, clusters, peak heights, cluster sizes, sample size, ...)
- Compare these parameters to the parameters of a meta-analysis constructed from HCP dataset
- Find new datasets (+ extract as many parameters as possible)
- Simulate new meta-analysis set based on these parameters
- Export different formats (t-maps, ES maps, SPM output, FSL output, peak coordinates)