Software
The software available below can be freely used for research and educational purposes.MCF: Module-constrained principal component analysis of brain connectivity
Given a set of connectivity (or weighted adjacency) matrices, modular connectivity factorization (MCF) (Hirayama et al., 2016) seeks to explain their variability by a small number of components where the corresponding basis or "eigenconnectivity" (Leonardi et al., 2013) patterns are explicitly constrained to have modular network structures. The method extends our previous orthogonal connectivity factorization (OCF) method (Hyvärinen, et al., 2016) so as to consider the variability in both inter- and intra-module connectivities and also to allow more than two modules.
The package also includes sample connectivity data.
Reference:
- J. Hirayama, A. Hyvärinen, V. Kiviniemi, M. Kawanabe and O. Yamashita. Characterizing variability of modular brain connectivity with constrained principal component analysis. PLoS ONE, 11(12): e0168180, 2016. [paper]
- A. Hyvärinen, J. Hirayama, V. Kiviniemi and M. Kawanabe. Orthogonal Connectivity Factorization: Interpretable decomposition of Variability in Correlation Matrices. Neural Computation, 28:445-484, 2016. [Matlab code & preprint]
- N. Leonardi, et al. Principal components of functional connectivity: a new approach to study dynamic brain connectivity during rest. Neuroimage. 83:937-950, 2013.
Acknowledgments: This work was partially supported by a contract with the National Institute of Information and Communications Technology entitled, “Development of network dynamics modeling methods for human brain data simulation systems,” Strategic International Collaborative Research Program (SICORP) from Japan Science and Technology Agency (JST), and KAKENHI 25730155, 15H02759 from Japan Society for the Promotion of Science (JSPS).
LCMM: Unified source separation and clustering analysis for EEG/MEG
A Matlab implementation of the statistical method proposed in (Hirayama et al., 2014,2015) for unsupervised EEG/MEG analysis. The method performs unsupervised clustering of sources' joint activity (coactivity) and simultaneously the blind separation of sources. The two stages of analysis are unified into learning a single probabilistic model, called Latent Coactivity Mixture Model (LCMM). See the references below for more technical details.
This software requires minFunc and Signal Processing Toolbox.
Reference:
- J. Hirayama, T. Ogawa and A. Hyvärinen. Unifying blind separation and clustering for resting-state EEG/MEG functional connectivity analysis. Neural Computation, 27(7):1373-1404, 2015. [paper]
- J. Hirayama, T. Ogawa and A. Hyvärinen. Simultaneous blind separation and clustering of coactivated EEG/MEG sources for analyzing spontaneous brain activity. The 36th Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC'14), pp.4932-4935, 2014.
Acknowledgments: This work was supported by a contract with the Ministry of Internal Affairs and Communications "Novel and Innovative R&D Making Use of Brain Structures" and by JSPS KAKENHI 25730155. The author was also partially supported by Strategic International Research Cooperative Program, Japan Science and Technology Agency (JST).