Comparing Structural Brain Connectivity by the Infinite Relational Model

The growing focus in neuroimaging on analyzing brain connectivity calls for powerful and reliable statistical modeling tools. We examine the Infinite Relational Model (IRM) as a tool to identify and compare structure in brain connectivity graphs by contrasting its performance on graphs from the same...

Full description

Saved in:
Bibliographic Details
Published in2013 International Workshop on Pattern Recognition in Neuroimaging pp. 50 - 53
Main Authors Ambrosen, Karen S., Herlau, Tue, Dyrby, Tim, Schmidt, Mikkel N., Morup, Morten
Format Conference Proceeding
LanguageEnglish
Published IEEE 01.06.2013
Subjects
Area measurement
Bayesian Methods
Brain models
Complex networks
Imaging
Mutual information
Neuroimaging
Predictive models
Relational Modelling
Structural Connectivity
Online AccessGet full text
DOI10.1109/PRNI.2013.22

Cover

More Information
Summary:The growing focus in neuroimaging on analyzing brain connectivity calls for powerful and reliable statistical modeling tools. We examine the Infinite Relational Model (IRM) as a tool to identify and compare structure in brain connectivity graphs by contrasting its performance on graphs from the same subject versus graphs from different subjects. The inferred structure is most consistent between graphs from the same subject, however, the model is able to predict links in graphs from different subjects on par with results within a subject. The framework proposed can be used as a statistical modeling tool for the identification of structure and quantification of similarity in graphs of brain connectivity in general.
DOI:10.1109/PRNI.2013.22