🧠 Excited to share our latest research in Imaging Neuroscience ! 🎓 "Rescuing Missing Data in Connectome-Based Predictive Modeling" 📈 Strategic imputation of missing data can significantly enhance prediction performance in brain-phenotype studies. 🔗direct.mit.edu/imag/article/d…

1/11 The poll has ended: 53% (scan time) vs 47% (sample size)! Here's our in-depth take on brain-wide association studies (with a 134-page supplement!): biorxiv.org/content/10.110… Led by Leon Ooi Csaba Orban, there are a few twists, so do read till the end! Our previous study ...

Excited to share that our paper "Data leakage inflates prediction performance in connectome-based machine learning models" is now out in Nature Communications nature.com/articles/s4146… (with Link Tejavibulya, @925452096Andy, Stephanie Noble (@sneuroble.bsky.social), and Dustin Scheinost )

New work by Chris Fields found increasing perceived discrimination from ages 10-11 to 13-14 among Black and Asian adolescents in the ABCD dataset. Notably, discrimination was highest among Black adolescents in advantaged, largely White neighborhoods. doi.org/10.1101/2024.0…

Can neuroimaging predictive models survive across diverse real-world data?🧠🌎We put models to the test across 3 large-scale unharmonized datasets with over 2,800 participants. See our new work in this special issue of Developmental Cognitive Neuroscience: sciencedirect.com/science/articl…