link to homepage

Institute of Neuroscience and Medicine

Navigation and service


A long-standing focus of our work is the development and application of meta-analytic approaches for the statistical aggregation of neuroimaging findings into consolidated information on functional neuroanatomy.

Picture of the topic "Meta Analysis"Schematic illustration of the steps of a meta-analysis: The coordinates reported in the individual experiments are extracted by creating a table reporting all x,y,z-coordinates. After modelling the spatial uncertainty associated with each individual coordinate, the resulting ALE-scores are assessed against a null distribution reflecting a random spatial association between experiments. Then, results are thresholded and corrected for multiple comparisons.

In the last two decades, neuroimaging has evolved to a standard method to investigate brain-behavior relationships and the pathophysiology of brain disorders. However, single neuroimaging data usually rely on small sample sizes and vary in experimental and analytical setup. In, addition, due to the large amount of published neuroimaging data, it is increasingly difficult to sift through the literature and distinguish spurious from replicable findings. Thus, there is a need to quantitatively consolidate effects across individual studies. One potent and popular approach for synthesizing the multitude of results in an unbiased fashion is to perform coordinate based meta-analysis. We evaluate, further optimize and apply the activation-likelihood estimation approach, which is one of the most widely used algorithms for neuroimaging meta-analysis. This method not only allows identifying those regions reliably associated with a particular process (e.g. working memory) or disease (e.g. depression), but the resulting regions can also be used as region of interest to guide analyses for newly acquired data. Another substantial implementation of ALE is the use for meta-analytic connectivity modelling (MACM), a powerful tool to investigate functional connectivity. MACM aims to derive networks consistently found to activate together with a specific seed region across experiments investigating different functions.
Thus, ALE meta-analyses allow us to capitalize on much of the published neuroimaging data, providing a quantitative summary of these data to answer specific topic-based research questions on brain-behavior associations and functional connectivity.
Additionally, a main focus of our research is to make the methods and results of meta-analyses openly available. Therefore, in collaboration with the university of San Antonio, GingerALE, a tool to perform ALE meta-analyses is provided, which is already widely used by researchers from all over the world. The results of meta-analyses can be shared via ANIMA, an open-source database, developed by our group, which allows researchers to upload their own and download other meta-analytic results.