Algebro-Topological Analysis of fMRI Connectomes Dynamics

This paper covers how algebro-topological analysis can be used for trying to understand global brain organization. The main goal was to understand if there are any patterns in human brain that can be mathematically explained and if so, can they be useful for, for example, diagnosing diseases. We were working with 138 resting state fMRI recording from the OASIS-3 open dataset, preprocessed by the laboratory of applied geometry and topology. This data gets preprocessed so we can build time series of dynamic simplices. After that, using dionysus2 framework, dynamic simplices were built in order to to distinguish zigzag persistent homology in them. The second goal was to get as many different visuals of our data as possible. As for now, no patterns were discovered. As a result, the python algorithm was created, but no patterns were found.