MEG optimal fingerprint functional connectome sub-networks

These files represent optimal fingerprinting functional connectome sub-networks from 90 regions of the Automated Anatomical Labeling atlas. Each sub-network contains 10/90 regions optimized on the fingerprint metrics of multi-session data of 43 subjects. The data is usable in the NumPy library with a recent version of Python.The search space is 5.72 trillion, and our data is 46,071 sub-networks in <code>optimalSubnetworksAll.npy</code> in form (46071, 10). In <code>optimalSubnetworksIndependent.npy</code> 24,197 sub-networks in form (24197, 10) are provided, independently sampled from unique optimization runs. <code>brainRegionsByIndex.npy</code>` in form (90,) stores a list of the names of the 90 AAL regions, where their list index corresponds to the values in <code>optimalSubnetworksAll.npy</code> and <code>optimalSubnetworksIndependent.npy</code>.<br><b>This data is generated using the code in the following Github link:</b>https://github.com/VasilesBalabanis/MEGSubnetworks<br><b>The bioRxiv Preprint link for the research study:</b>https://www.biorxiv.org/content/10.1101/2024.06.23.599587v1<b>Citation:</b>Please cite this data as: V.Balabanis, J.Zhang, X.Xie, S.Yang. (2024). Robust sub-network fingerprints of brief signals in the MEG functional connectome for single-patient classification. Figshare. DOI: 10.6084/m9.figshare.26130376