Dear All,
I am relatively new to the field and psychologist by training.
I was just wondering what the typical workflow of the mindboggle ouput for Laplace Beltrami spectra or Zernicke Moments would be? How would you proceed with the output to get emaningful results (by means of group comparisons or regression analyses)?
Thanks for your answers
Mindboggle outputs include tables with many features, some features per cortical surface vertex and some features per collection of vertices. For each collection of vertices such as a cortical region, Mindboggle computes statistical summary measures, Laplace-Beltrami spectral coefficients, and Zernike moments. People usually use these features in machine learning or statistical analyses of the morphological properties across brains or for a given brain over time. Examples where people have applied different algorithms to analyze brain shape data include the Alzheimer’s big data challenge (https://www.synapse.org/#!Synapse:syn2290704/wiki/60828) and in the primary reference for Mindboggle (http://journals.plos.org/ploscompbiol/article?id=10.1371/journal.pcbi.1005350).
Martin Reuter grup, from MIT have two really interesting papers (and available code here!) where they explore several statisticals approaches to use the Laplace-Beltrami spectral coefficients. The refs:
[ 1 ] Wachinger C; et al. Whole-brain analysis reveals increased neuroanatomical asymmetries in dementia for hippocampus and amygdala. Brain. doi: 10.1093/brain/aww243
[ 2 ] Wachinger C; et al. BrainPrint: a discriminative characterization of brain morphology. NeuroImge. doi: 10.1016/j.neuroimage.2015.01.032
MANAGED BY INCF