Abstract #3550

Online Learning for Real Time fMRI Classification

Hao Xu¹, Yongxin Taylor Xi¹, Ray Lee², Peter J. Ramadge¹

¹Electrical Engineering, Princeton University, Princeton, NJ, United States; ²Princeton Neuroscience Institute, Princeton University, Princeton, NJ, United States

We propose a real-time conjugate gradient(rtCG) algorithm as an efficient solution for online real-time fMRI(rtfMRI) system. rtCG leans to classify brain states as data is being collected. It has a close connection with well-established partial least squares(PLS) algorithm when applied to the quadratic problem of interest. Real fMRI data tests show that rtCG can process high dimensional fMRI data within one TR and reach high prediction accuracy.