Souvik Phadikar
Research Scientist I
TReNDS Center & Georgia State University
I am a Scientist I at the TReNDS Center. My research focuses on multimodal brain imaging, with an emphasis on EEG-fMRI fusion methods, independent component analysis (ICA), and psychiatric neuroimaging. Different brain imaging modalities each capture distinct aspects of brain activity, and analyzing them separately leaves information on the table. I develop methods that jointly analyze multiple imaging sources to extract results that are both interpretable and biologically grounded. The longer-term goal is to study major depressive disorder and identify multimodal brain patterns that distinguish its subtypes.
I have built a series of joint ICA-based fusion frameworks: group joint ICA (gjICA), Dynamic group joint ICA (Dyna-gjICA), and parallel multilink group joint ICA (pmljICA).
Before joining TReNDS, I was a postdoctoral researcher at the University of Wisconsin-Madison, working on automated seizure prediction for hospitalized patients. We developed a model that predicts seizures up to four hours before onset.
During my PhD at NIT Silchar, I worked across signal processing, developing methods for automatic EEG artifact detection and removal, and building brain-computer interface systems for motor imagery, silent speech, and emotion classification.