In our recent article, Ajith et al. compare three deep learning models for predicting brain health using unimodal and multimodal strategies doi.org/10.52294/001...@vcalhoun.bsky.social@fmri-today.bsky.social
Also be sure to check out the GIFT BIDSapp which includes the Neuromark templates and the fully automated spatially constrained ICA pipeline (github.com/trendscenter...)!
Contribute to trendscenter/gift-bids development by creating an account on GitHub.
Templates, including our 105 component multi-scale template (pubmed.ncbi.nlm.nih.gov/37787573/trendscenter.org/datatrendscenter.org/software/gift).
With the new templates, the NeuroMark framework allows us to perform age-specific adaptations and to capture features adaptable to each modality, therefore facilitating biomarker identification across brain health and disorder.
Our results suggest remarkably high similarity of the resulting four-dimensional templates (age × 3D brain), demonstrating the generalizability of intrinsic brain networks across ages.
We introduce two significant expansions of our NeuroMark templates (NeuroMark 3.0) first by generating replicable ICA of fMRI templates for infants, adolescents, and aging cohorts, and second by incorporating structural MRI and diffusion MRI modalities (sciencedirect.com/science/arti...).
You should add GIFT and FIT to this page! ;-) trendscenter.org/software/gifttrendscenter.org/software/fittrendscenter.org/software .
Some of the upcoming exciting news: backend redesign and refactoring leading to a 10x reduction in codebase! The new COINSTAC front-end design will also knock your socks off ;) (Coming soon!) Here's to many more successful gatherings!
Our COINSTAC (https//coinstac.org/) on site meeting/lunch....nice fall day in Atlanta...a centralized meal for a federated team. ;-)
yeah, also helps prevent the 'perfect is the enemy of the good' paper that never gets published. My first paper was like this, I still have it, it's really good, but never published. ;-)